Journal of Archaeological Science: Reports 34 (2020) 102637 ELSEVIER Contents lists available at ScienceDirect Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep Neolithic land-use, landscape development, and environmental dynamics in the Carpathian Basin Michael Kempf a'h* a Department of Archaeology and Museology, Masaryk University, Brno, Czech Republic, Arne Nováka 1, Brno 60200, Czech Republic b Physical Geography, Institute of Environmental Social Science and Geography, Faculty of Environment and Natural Resources, University of Freiburg, Schreiberstr. 20, 79085 Freiburg Germany (D ARTICLE INFO ABSTRACT Keywords: Holocene environmental dynamics and the onset of early subsistence farming during the Neolithic period have Holocene jecj to major surface modification and landscape transformation in the Carpathian Basin. In this context, Neolithic Neolithisation settlements and agricultural activities were supposed to be located on Chernozem soil patches, which originated Hungary from loess-covered surfaces of the uneroded Pleistocene and early Holocene palaeolevees. Chernozem soil dis- Chernozem Soil development tnbution is seen as an important precondition of agricultural expansion. However, Chernozem soil genesis and Spatial modelling ^e anthropogenic modification of soil organic matter and Black Carbon (BC) content from clearing and vege- Landscape archaeology tation burning are not yet fully understood and there is increasing evidence for the active role of human land- scape interaction in the process of Chernozem development. Consequently, Neolithic land-use would not have been necessarily linked to Chernozem but rather triggered its development from alluvial and meadow soils through intensified surface transformation. This article applies a GIS-based multivariate surface analysis and a statistical evaluation to 49 Neolithic sites to track environmental location factors, soil preferences, and potential land-use strategies in Neolithic Hungary. The combination of remotely sensed surface data, environmental GIS-attributes, quantitative statistics, and archaeological datasets reveals site-location parameters during the Early to the Late Neolithic and critically discusses Chernozem soil development and utilization during agricultural transformations across Europe. 1. Introduction Various interdisciplinary studies on Early to Late Neolithic mobility and migration, land-use, and settlement development have been carried out recently, which result in detailed knowledge about cultural and social processes in prehistoric Hungary (Domboroczki and Raczky, 2010; Giblin et al., 2013; Gronenborn, 1999; Haak et al., 2005; Salisbury, 2013; Whittle, 2007). The rapid spread of early farming techniques and the uniformity of the archaeological material culture (Banffy, 2019; Bille and S0rensen, 2019; Bonsall et al., 2002) indicate intense mobility, communication, and exchange patterns on a considerable short-term temporal scale (Banffy, 2019; Haak et al., 2015; Szecsenyi-Nagy et al., 2015). In that period, an early domestic subsistence economy and a highly diverse mosaic of permanent settlements and transient camps on loess-covered areas and in close connection to running fresh water has been assumed (Hedges et al., 2013; Sielmann, 1972; Whittle, 2007). This Neolithic landscape, however, developed from dynamic riverbed fluctuations and annual flooding, which created a mosaic landscape pattern characterized by backswamps and non-permanent oxbow lakes in reactivated Holocene palaeochannels (Gillings, 1995, 2007). So-called avulsion events (abandonment of the established river channel in favor of a new channel in the adjacent floodplain (Heyvaert and an and Walstra, 2016)), which occurred prior to the transformation of the anastomosing river system, frequently led to the establishment of new channel systems. Eventually, the palaeolevees remained uneroded after the stabilization of the run-off behavior and the development of a meandering streamflow character (Magyari, 2011). In this setting, the former palaeochannels were still periodically flooded during heavy rainfall events, which triggered the deposition of fine-grained clayey material and the subsequent formation of waterlogged soils that reinforced the flooding vulnerability and the development of oxbow-shaped wetlands and swamps. It is highly debatable, whether these soils were agriculturally exploited or if we lack traces because human activity is * Address: Department of Archaeology and Museology, Masaryk University, Brno, Czech Republic, Arne Nováka 1, Brno 60200, Czech Republic E-mail addresses: kempf@phil.muni.cz, michael.kempf@archaeologie.uni-freiburg.de. https://doi.Org/10.1016/j.jasrep.2020.102637 Received 16 July 2020; Received in revised form 21 September 2020; Accepted 24 September 2020 Available online 26 October 2020 2352-409X/© 2020 Elsevier Ltd. All rights reserved. M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 buried under the alluvial deposits (Sherratt, 1983). Alternatively, continuous occupation of the elevated loess-covered areas with thick and fertile Chernozem coverage has been assumed (Garnett, 1945; Hedges et al., 2013; Kreuz, 2007; Sielmann, 1972). Recently, however, Eckmeier et al. (2007) emphasized that buried Neolithic soils do not necessarily need to be connected to in situ developed fertile Chernozem soils but can carry a significant amount of charred material with up to 70% total organic carbon - a soil-charcoal mixture that gives the impression of a relict Chernozem soil (Acksel et al., 2017; Eckmeier et al., 2007; Schmidt et al., 1999; Schmidt and Noack, 2000). This raises the question to what extent thick and fertile Chernozems were already developed in the Neolithic or if their development was connected to local steppe vegetation, forest coverage, climate oscillation, and human surface modification, land-use, and deforestation activities (Barczi et al., 2006; Gerlach and Eckmeier, 2012; Lorz and Saile, 2011; Preston and Schmidt, 2006; Schmidt et al., 1999; Schmidt and Noack, 2000; Starling, 1983; Strouhalová et al., 2019; Vysloužilová et al., 2015). This article addresses multi-layered environmental surface data to estimate Holocene landscape development and the impact of Neolithic surface modifications on soil development in Hungary (Fig. 1). A comprehensive understanding of ecological and palaeo-environmental feedbacks supports the hypothesis that fertile and extensive Chernozem coverage could be strongly connected to Neolithic agricultural exploitation, forest clearing, and vegetation change. 2. Material and methods The sample, which is used in this article is based on the database of the research project Population history of the Carpathian Basin in the Neolithic Age and its influence on settlement in Central Europe, which was funded by the German Research Foundation (DFG, grant number Al 287/10-1) and led by Eszter Banffy and Kurt Alt. Through the integration of various methods from archaeology, anthropology, molecular genetics, and biogeochemistry, it was possible to visualize the structure and dynamics of settlement processes and population development in the Carpathian Basin in their spatio-temporal depth (Banffy, 2019; Haak et al., 2015; Lipson et al., 2017; Mathieson et al., 2015; Szecsenyi-Nagy, 2015; Szecsenyi-Nagy et al., 2015). One sub-study focused on strontium and oxygen isotope analyses led on Neolithic populations to track past human mobility patterns (Depaermentier et al., 2020a, 2020b). It covers the Hungarian part of the Carpathian Basin and is based on the same sample that has already been used for nitrogen and carbon stable isotope as well as aDNA analyses (Szecsenyi-Nagy, 2015; Szecsenyi-Nagy et al., 2015). In this context, a comprehensive Holocene landscape evaluation was carried out to address site location hypothesis and potential land-use strategies in relation to strontium and oxygen stable isotope analysis in the Carpathian Basin (Depaermentier et al. 2020). Consequently, this paper includes 49 Neolithic sites, which were also investigated for mobility patterns on the site-specific to the micro-regional scale (Depaermentier et al., 2020; Kempf et al., 2020) (see Fig. 1 and Table 1. for location and chronological periods). 2.1. Archaeological sites The western part of the Carpathian Basin played a major role in Europe during the Neolithic period. Transdanubia and the southern lowlands were the most northerly distribution area of the so-called Starcevo-Kbrbs-Cri§ cultural complex (see Table 1). In the central part 21.0 Neolithic sites in Hungary 48.0 50 100 km -1-1-1-1-1— 16.5 18.0 19.5 21.0 22.5 Fig. 1. Neolithic site distribution. 49 sites in Hungary were analyzed for site-catchment properties and potential land-use strategies. 48.0 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Table 1 Site names, abbreviations, and chronological periods of the sites used in this article. The geological and pedological information of each site is given using the major soil type classification of Hungary (see Tab. 2 for the correlation with the WRB system) and the geological EGDI datasets (see Fig. 4). The classification of the sites is based on the database of the DFG-project. Some sites comprise multifold chronological periods and the database also includes Chalcolithic occupation, which was additionally modelled for the sake of completeness (EN = Early Neolithic, MN = Middle Neolithic, LN = Late Neolithic, CHAL = Chalcolithic). Site name id Culture Period lat Ion Soil units Geological units Abony 60. lh. ABO Tiszapolgär, Szakälhät MN/ 47.19 20 sand soil diamicton CHAL Alsónyék elkerülö 2. lh. ALE Sopot MN 46.2 18.72 meadow soil diamicton Balatonszemes Bagódomb BAB TLBK MN 46.81 17.78 meadow soil silt Bátaszák-Lajvér BAL Lengyel II LN 46.2 18.7 meadow soil diamicton Bátaszék-Mérnóki telep BAM Starcevo EN 46.21 18.71 meadow soil diamicton Berettyóújfalu-Nagy-Bócs diilö BENA Esztär, Koros, Hungarian Conquest Period EN/MN 47.23 21.53 meadow soil diamicton Bicske Galagonyás BICS Sopot MN 47.49 18.66 salt affected diamicton Borjád Kenderföldek BORK Lengyel HI (?) LN 45.94 18.47 meadow soil silt Bölcske Gyürüsvölgy M3-TO 14. lh. BÖVÖ TLBK - Zseliz MN 46.74 18.96 sand soil diamicton Budakeszi 8. lh. Szölöskert-Tangazdasag BUD TLBK MN 47.51 18.93 brown forest impure soil limestone Cegléd CEG Szakälhät MN 47.17 19.8 meadow soil silt Cegléd Ipari park CGIP Szakälhät MN 47.2 19.79 meadow soil silt Lanycsók Csata alja CSAT Lengyel LN 47.51 18.62 brown forest silt Deszk Ordos DEOR Tisza LN 46.21 20.23 peat soil diamicton Ebes-Zsong-völgy EBVÖ Esztär MN 47.48 21.5 Chernozem silt Fajsz FAGA Sopot MN 46.42 18.92 alluvial soil diamicton Felsöörs-Barokert FEB Lengyel II LN 47.02 17.96 brown forest limestone Garadna- Elkerülö 2.1h GAEL Bükk, ALBK MN 48.41 21.17 soil brown forest diamicton Hajdunanas-Eszlari ut, M3-46 HAJE ALBK MN 47.85 21.43 Chernozem diamicton Harta-Gátórház HARG TLBK MN 46.7 19.03 Chernozem diamicton Hejökürt-Lidl logisztikai központ HELI Tiszadob MN 47.87 21.01 meadow soil diamicton Hódmezóvásárhely-Gorzsa V. lh HOGO Baden CHAL 46.37 20.43 meadow soil diamicton Homokbänya Hódmezóvásárhely Kotacpart HOKO Koros EN 46.42 20.33 meadow soil diamicton Keszthely-Fenekpuszta KEFP Balaton-Lasinja CHAL 46.71 17.24 meadow soil org. rich Pusztaszentegyházi dülö sediment HMV Kökenydomb KOKE Tisza LN 46.42 20.33 meadow soil diamicton Kompolt-Kígyós-ér KOKI ALBK MN 47.73 20.24 meadow soil tuff-breccia Kóny 85 Enese KON TLBK, Latest Lengyel/Balaton Lasinja, late MN 47.64 17.36 meadow soil diamicton TLBK Zseliz, Lengyel III Lanycsók Gata Csatola LCGS Starcevo EN 46.01 18.62 meadow soil silt Lanycsók Csata alja M6- Starcevo, Latest Baden/Vucedol, Balaton Lasinja EN/ 46 18.58 Chernozem diamicton 116 CHAL Mezökeresztes-Cethaloni MECE ALBK 2-3 MN 47.8 20.69 meadow soil diamicton Mezözombor - Közsegi temetö MEKÖ ALBK, Bükk, Tiszadob MN 48.14 21.26 meadow soil diamicton Mórágy Tüzködomb Bl MORT Lengyel LN 46.21 18.65 brown forest silt Nemesvanios-Kapsa NEK Sopot MN 47.05 17.88 brown forest silt Polgär-Piocäsi-Dülö POPI ALBK MN 47.86 21.12 salt affected diamicton Pusztataksony Ledence PULE Tisza, Szakälhät, Tiszapolgär, ALBK MN/LN 47.45 20.51 soil sand soil diamicton Sajoszentpeter-vasúti örhaz SAVÖ Bükk, ALBK MN 48.22 20.71 skeletal soil diamicton Szederkeny-Kukorica-dülö 95. lh. SEKU Vinca MN 46 18.45 meadow soil silt no data SZEG Tisza LN 46.58 20.22 alluvial soil diamicton Szemely-Hegyes M60/83. lh. SZEH Sopot, TLBK Mn 46.03 18.32 brown forest silt Tiszabura Bonishat TIBO ALBK MN 47.45 20.46 sand soil diamicton Tiszaszölös - Domaháza-puszta, Réti- TIDO Szakälhät, Koros, Szakälhät or early Tisza, ALBK EN/MN 47.56 20.72 alluvial soil diamicton dííló Tiszalók Hajnalos TIHA no data MN 48.02 21.37 meadow soil silt Tiszadob-Okenéz TISO Tiszadob/Bükk MN 48 21.17 alluvial soil diamicton Tiszafóldvár Téglagyár TITE Szakälhät MN 46.97 20.25 meadow soil diamicton Tolna-Mózs TO 026 TOLM Balaton-Lasinja, TLBK MN/ 46.41 18.74 Chernozem silt CHAL Debrecen Tócópart Erdóalja TOPE Esztär MN 47.52 21.58 meadow soil silt Versend-Gilencsa VEGI Vinca MN 46 18.51 Chernozem mudstone Veszprém-Jutasi-Munkacsy út VEJ Lengyel II, Balaton-Lasinja LN/ 47.1 17.91 brown forest silt CHAL soil Visonta VISO ALBK MN 47.78 20.03 meadow soil diamicton 3 M. Kempf of Transdanubia, the first Neolithic arable farming culture in Central Europe (Linearbandkeramik, LBK, 5500 to 4800 cal BC) developed around 5500 cal BC - probably in strong interconnection with hunter-gatherer groups (Banffy, 2004, 2006; Banffy and Bognar-Kutzian, 2007; Banffy et al., 2008; Eichmann et al., 2010). Eventually, the Transdanubian LBK (TLBK) spread rapidly across Central Europe (Szecsenyi-Nagy et al., 2014) and existed alongside the Central European LBK (Oross and Banffy, 2009). Around 5300/5200 cal BC, the so-called Vinca culture influenced southwestern Hungary (Jakucs et al., 2016). After 5000 cal BC, the so-called Sopot culture alternated the LBK, which introduced new cultural characteristics at the onset of the Late Neolithic in Transdanubia. From 4900/4800 cal BC onwards, the Sopot impulses triggered the formation of the large Late Neolithic Lengyel cultural complex, which represented imprints from both, the LBK and the Sopot material culture. The Lengyel period was characteristic across Europe (Stadler and Ruttkay, 2007) and in its late phase reached into the Chalcolithic period around 4300 cal. BC. Finally, the so-called Balaton-Lasinja culture marked the end of the transition from the Late Neolithic and Early Chalcolithic to the Middle Copper Age. In the lowlands, the Hungarian Early Neolithic is characterized by the so-called Koros culture, which spread over the northern Balkans from the Maros region of the southern Alfold after 6000 cal BC along the rivers Tisza, Koros, and Berettyo (Domboroczki and Raczky, 2010; Raczky et al., 2005; Whittle et al., 2002). The Middle Neolithic of the Hungarian Plain is mostly dominated by the Alfold Linear Pottery culture (ALBK), which formed around 5600-5400 cal BC. The ALBK spread from its origin over the entire Hungarian lowlands (Oross and Banffy, 2009). Around 5300 cal BC, a differentiation process began, in the course of which several regional groups emerged (Hertelendi et al., 1995). Along the lower and middle parts of the river Tisza, the so-called Szakalhdt culture (Kalicz and Makkay, 1977) was one of the most widespread group, which occupied a large part of the former distribution area of the Koros culture. The distribution area of the Bukk group partly overlapped with that of the Zseliz, Esztar and Szakalhdt groups (Raczky and Anders, 2009). The Szakalhat culture also maintained contacts with the Middle Neolithic so-called 'note-head' and Zseliz groups in Transdanubia and with the Vinca culture in the northern Balkans (Kalicz and Makkay, 1977). The northern Alfold including the Bukk and Matra mountains were the distribution area of the so-called Tiszadob group, which first appeared around 5300 cal BC, and the subsequent Bukk group (Csengeri, 2010; Kalicz and Makkay, 1977). Around 5100 cal BC the Tisza culture appeared in southern Alfold (Hertelendi and Horvath, 1992; Kalicz and Raczky, 1987) and spread to the central and northern Alfold (Kalicz, 1994; Kalicz and Raczky, 1987). In the statistical analysis, which was carried out using the entire Neolithic and Chalcolithic data, both chronological periods were integrated for the sake of completeness. The Neolithic sites were further differentiated into Early Neolithic (EN), Middle Neolithic (MN), and Late Neolithic (LN) following the Hungarian chronological periods (see Table 1). A few sites show more than one cultural and chronological occupation. 2.2. Climatic conditions Hungary is dominated by a moderate continental climate with marine and Maritime influences (Acs, Breuer and Skarbit, 2015; Demeny et al., 2013). The topographical conditions and the large-scale of the Carpathian Basin cause an increase in continentality towards the lowlands with lowest precipitation rates in the center of the plain (Fig. 3) (Jakab et al., 2009). During the Holocene, the harsh climatic conditions were reinforced by desiccating winds that limited tree growth and favored the accumulation of sandy deposits (Gardner, 2002; Novothny et al., 2010). That led to an Early Holocene steppe/forest-steppe vegetation cover followed by a rapid forest decline and niche forest habitats surrounded by an extensive steppe plain (Gardner, 2002; Magyari et al., 2010; Uj et al., 2016; Willis et al., 1997, 2000). The Boreal in the Great Journal of Archaeological Science: Reports 34 (2020) 102637 Hungarian Plain was characterized by grassland surfaces and a regional climate that was drier and warmer than today (Hertelendi et al., 1992; Magyari et al., 2010; Magyari, 2011). Magyari (2011) and Magyari et al. (2010) see a tendency towards drier conditions after 8.4 ka BP that were accompanied by decreasing lake levels and increased fire outbreaks as reported from micro-charcoal dispersal (Magyari et al., 2010; Magyari, 2011). The subsequent Atlantic phase is assumed to have experienced a climatic optimum approximately 9 k to 5 k cal BP (Hedges et al., 2013; Hertelendi et al., 1992; Magyari et al., 2001). 2.3. Sedimentohgical conditions Large parts of the Great Hungarian Plain (Alfold) are covered with Quaternary sediments, which are composed of gravel, sand, silt, and clay (Kercsmar et al., 2015). The Upper Pleistocene loess is characteristic for the hilly margins of the plain, the Mezofold west of the Danube, and mainly the alluvial fans of the basins. The origin, thickness, and age of the loess deposits show local discontinuities (Varga, 2011). The deposits can alternate with sandy layers and are frequently disturbed by palae-osoil horizons, which developed during warmer and wetter interglacial periods (Kovacs et al., 2013). Massive sand deposits occur mainly in the Danube-Tisza Interfluve and the Kiskunsag east of the Danube (Kercsmar et al., 2015). The Danube-Tisza Interfluve (hereinafter termed DTI) is the largest sand-covered area in Hungary (Fig. 2, Fig. 4). The wind-blown sediments derived mostly from the palaeo-alluvial fans of the rivers and are connected to Quaternary landscape formations under increased wind speed and activity (Ladanyi et al., 2015; Novothny et al., 2010). Finally, the most significant aeolian deposition occurred in the Upper Plen-iglacial and was again reworked during the cold and dry phases during the Older and Younger Dryas (Nyari et al., 2014). In the Early Holocene, the gradual increase of the mean temperature reached today's characteristics with a July value of 22 °C as reported from mollusc faunal analysis by Siimegi et al. (2015). The data samples were taken from carbonate rich lakes that formed at the DTI at the end of the Pleistocene and Early Holocene when the Danube accumulated a huge alluvial fan (Knipl and Siimegi, 2012). Aeolian redistribution and relocation of the sandy deposits formed small-scale basins where water was trapped after flooding events. The mineral mixture of the lake water and the carbon dioxide entrapment in combination with the saline and alkaline salt content led to the sedimentation of dolomite mud during the Neolithic (Siimegi et al., 2015). The eastern part of the Carpathian Basin is dominated by hydrologic channel shifts and floodplain dynamics of the late Pleistocene/Holocene river Tisza (Kasse et al., 2010; Kiss et al., 2015; Moskal-del Hoyo et al., 2018; Timar et al., 2005), the river Koros (Petrovszki and Timar, 2010), and the river Maros (Kiss et al., 2014). The sand-covered DTI experienced alterations of fluvial deposits from the Danube, palaeosoil development after the west drift of the river, and depositional activity superimposed on palaeosoils (Nyari et al., 2014). 2.4. Soil conditions and textures According to Laborczi et al. (2016), soil texture information is an essential data source for modelling climatic, hydrological, and agricultural conditions. Surface cover transformations and degradations caused by wind erosion or compaction strongly depend on local topsoil textures (Birkas et al., 2004; Laborczi et al., 2016; Varallay, 1989). For this reason, the research group among Laszlo Pasztor1 converted large datasets of Hungarian soil survey information into digital formats that are organized in spatial soil information systems (SSIS) (Laborczi et al., 1 Department of Environmental Informatics, Institute for Soil Sciences and Agricultural Chemistry, Centre for Agricultural Research, Hungarian Academy of Sciences, Budapest, Hungary. 4 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 16.5 18.0 _I_ 19.5 21.0 22.5 Mean annual total precipitation in Hungary [1970-2000] 24.0 ■ Rivers and lakes Budapest Precipitation [mm] ■ 600 650 700 750 50 100 km —i— 24.0 Fig. 3. Modelled mean annual total precipitation in Hungary for the reference period 1970 - 2000. The continental climate with dry conditions and annual total precipitation amounts around 500 mm generates steppe vegetation in large parts of the Great Hungarian Plain. The moisture regime varies from moderately moist (Ukrainian border) to moderately dry and dry (central part of the Tisza plain) (Acs, Breuer and Skarbit, 2015). Climate data based on worldclim.org (Hijmans et ak, 2005). 2019; Pasztor et al., 2012, 2016, 2014a, 2014b, 2015a, 2015b, 2002; Waltner et al., 2018). Finally, soil maps of different spatial resolution were generated that allow for the evaluation of soil textures and types as well as degree of degradation, salinization, and erosion (Fig. 5 and Fig. 6) (Pasztor et al., 2012, 2016; Schofield et al., 2001). Additional information on wetland dynamics, floodplain development and flooding vulnerability can be obtained from the data (Bozan et al., 2018; Pasztor et al., 2015a, 2015b). 5 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 16.5 18.0 21.0 _I_ Geological units in Hungary Geological units / landcover I I andesite I I olivine basalt I I diamictite I I gabbroic rock I I granite I I tuff-breccia I I impure dolomite n diamicton I I impure limestone n sand I I limestone □ silt n mudstone n no data □ organic rich sediment □ sandstone 0 50 100 km Fig. 4. Geological units in Hungary. The floodplains of the river Danube and the Tisza river-system are characterized by Holocene unsorted or poorly sorted, clastic sediments with a wide range of particle sizes. The DTI is mainly composed of sandy deposits. Large parts of the Carpathian basin and the hilly margins are covered with silty deposits and silty mud. The EGDI dataset refers to the units as follows: Diamicton = unsorted/poorly sorted clastic sediments, Holocene; silt = mud that consists of greater than 50% silt-sized grains, Pleistocene; sand = clastic sediments (<30% gravel), Holocene; organic rich sediment = greater than 50% organic material, Holocene; limestone = Triassic carbonates; impure limestone = Eocene impure carbonate sediments; tuff breccia = Miocene pyroclastic rock. Topsoil texture in Hungary USDA soil texture classes [depth 0-5 cm] ^\ sand □ clay loam ~| loamy sand ~| loam ^\ sandy loam □ silt ^\ sandy clay loam | silty clay □ silty loam □ clay ~| silty clay loam ~| no data 0 50 100 km (modified by Michael Kempf, 2019; based on Laborczi, A.; Szatmari, G.; Takacs, K. & Pasztor, L, 2015) 16.5 18.0 19.5 21.0 22.5 24.0 Fig. 5. Topsoil texture in Hungary based on the supplementary data of A. Laborczi, G. Szatmari, K. Takacs and L. Pasztor (2015) in (Laborczi et al., 2016). The most used soil classification system is the World Reference Base for Soil Resources (WRB). In this context, the Hungarian soil classification system was first correlated to the WRB classification by Micheli et al (2006) and Krasilnikov et al. (2009) (Balla, Novak and Zichar, 2016; Krasilnikov et al., 2009; Micheli et al., 2006). According to Krasilnikov et al. (2009), the system originally is based on four types (main type, soil 6 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 21.00 22.50 24.00 Soil units in Hungary Landcover A. Neolithic sites I I River system I I No data 50 Soil units I I Alluvial soils I I Meadow soils I Peat soils I I Brown forest soils I I Skeletal soils I I Chernozem I I Sand soils I I Lithomorphic soils I I Salt affected soils 100 km □ i Data provided by LaszIO Pasztor; see Pasztor, L. eta!. 2018, redrawn (c) Michael Kempf, 2019 Fig 6. Neolithic study sites modelled on soil units in Hungary. The sites follow a spatial pattern with preferred occupation of meadow soils and Chernozems that are frequently distributed in the central part of the Carpathian Basin and along the broad floodplain patches of the major river (soil data kindly provided by Laszlo Pasztor). type, subtype, and varieties), which can be correlated to the WRB as shown in Table 2. Hungarian soils cover an area of approximately 9.3 Mha of which 6.4 Mha are agriculturally used and 4.5 Mha are arable land. Over 40% show degradation processes and over 20% are compacted due to human activity (Birkas et al., 2004; Schofield, Thomas and Kirkby, 2001; Szi-lassi et al., 2006). Hungary shows very heterogenous soil compositions according to the manifold geological basements and sedimental fluvial infills. In the northern mountainous areas and on eroded slopes with mainly consolidated limestone, dolomite, basalt, and andesite, lithomorphic soils are present. Although these soils can show very particular physical and chemical aspects, they share similar vegetation (forest-) coverage and are grouped together (Dobos et al., 2000). Further south and in the western, more humid part of Hungary, Brown Forest soils (Haplustalfs) developed that show significant clay illuviation. Descending from the mountains, the clay content decreases and the transition zone to the lowlands is characterized by Chernozem-Brown soils Table 2 Correlation of the Hungarian soil classification and the World Reference Base for Soil Resources (WRB) based on Micheli et al (2006) and Krasilnikov et al. (2009). Hungarian soil classification WRB classification Skeletal soil Leptosols/Arenosols/Regosols/Calcisols Lithomorphic soils Leptosols/Regosols/Phaeozems/Cambisols Brown forest soils Cambisols/Luvisols/Umbrisols Chernozems Chernozems/Phaeozems/Kastenozems/Vertisols Salt-affected soils Solonchaks/Solonetz/Verti sols/Chernozems Meadow soils Humic Gleysols/Mollic Gleysols/Gleyic Phaeosems/ Gleyic Chernozems Sand soils Arenosols/Cambisols/Lamelli-Arenic Luvisols Peat soils Histosols Alluvial soils Fluvisols (Haplustolls and Hapludolls). The loess-covered plains are mainly dominated by Chernozems (Pachic and Typic Claciustolls) with distinct hy-dromorphic conditions in depressions (Aquic Calciustolls and Haplustolls) (Dobos et al., 2000). Over the last full glacial period, the central plain suffered from very dry climatic conditions with low precipitation rates and saline soils already started to develop accordingly - triggering a saline-tolerant vegetation that persisted throughout the Holocene (Magyari, 2011). Saline groundwater reached the soil surface and the annual evaporation was higher than the annual precipitation rate. Consequently, salt-affected soils developed (Schofield et al., 2001; Toth et al., 2001). In the DTI, blown sand (Typic Ustipsamments and Quartzipsamments) is frequently abundant (Dobos et al., 2000). However, a significant salt impact from the closed off Para Tethys Sea sedimentation regime with saline marine and lacustrian infills can also be detected in the DTI and especially in the Tisza plain (Madl-Szonyi et al., 2008; Toth et al., 2001). Quaternary loess originating from the Russian Puszta further strengthened the saline signal in the groundwater and the fine-grained soils (Solonchak soils) (Fig. 7) (Schofield et al., 2001). 2.5. Hydrologic systems and hazards Modern Hungary is located at the western border of the Eurasian Steppe Belt (Molnar et al., 2012; Pinke and Lovei, 2017). The country experienced not only a highly dynamic environmental and cultural history but also a turbulent recent transformation with political and economic developments and upheavals (Banffy, 2019; Bentley et al., 2012; Bickle et al., 2017; Schulze, 2000). Crop production became increasingly important in Hungarian 19th and 20th century modernization what led to massive impacts on the environmental settings of the country through canalization and drainage activity of the broad wetlands, wind driven soil erosion, and local agricultural overstraining (Pinke and Lovei, 2017; Uj et al., 2016; Ujhazy and Biro, 2018; van der 7 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Fig. 7. Salt-affected soils (purple) and the potential floodplains of the main river systems in Hungary (blue). Saline soils are mostly located in the Tisza floodplain and the DTI. Neolithic sites (red triangles) are situated in more than 250 m distance to the saline soils that show only a poor agricultural potential (one site (CEG) is situated on a saline soil patch of 800 m diameter). Soil data is based on (Pasztor et al., 2018); the potential floodplain is projected from Corine Riparian data. Zee et al., 2017). Consequently, Pleistocene and Early Holocene river channels and systems in Hungary vary greatly from the modern modified hydrologie streamflows. The hydrological development of the river Danube is strongly connected to the sedimentation regime during the late Pleistocene and the Early Holocene. The eastern part of the broad alluvial plain is bordered by a slightly elevated plateau-like palaeo-al-luvial fan that was first deposited by the river and then cut several times during flooding events and avulsion activity, which caused relocation of the alluvial deposits (Knipl and Siimegi, 2012; Ujházy et al., 2003). Finally, the DTI was created that mainly consist of relocated alluvial sands with local loess cover. The modern river Tisza is the second largest river in Hungary and one of the main tributaries to the river Danube (Kasse et al., 2010; Nagy et al., 2010). Subsidence of the Great Hungarian Plain controlled the drainage and hydrologie systems through basin development. Consequently, large-scale alluvial fans with fragmented palaeochannels were created by the river. Quaternary fluvial deposits mainly consist of finegrained clay, silt, and sand with a seismic thickness of up to 700 m (Kasse et al., 2010; Kiss et al., 2015; Moskal-del Hoyo et al., 2018; Nádor et al., 2011; Timár et al., 2005). Prior to massive channeling and regulations in the 19th century, large parts of the floodplain were periodically inundated (Bezdán, 2010; Guida et al., 2015) and the riverbed changed frequently (Kasse et al., 2010; Sipos et al., 2016). Holocene floodplain dynamics, as reported by Kasse et al. (2010), show a continuous trend of riverbed relocation and avulsion of more than 4 km in distance. The discharge of the fluvial system fluctuated due to climate variability: during humid and rather warm conditions, riverbed dynamics are considered to follow a meandering pattern, while the subsequent, short-term dry and cold phase triggered a braided river system of the Tisza-tributaries due to sparse vegetation coverage (Kiss et al., 2015). The Tisza discharge levels increased in the Atlantic Phase due to humid conditions what supported the development of large meanders in the Upper Tisza region and a general discharge volume of 5520 m3/s -20 times greater than the Boreal value and 10 times greater than the modern ratio (Kiss et al., 2015). The high flooding vulnerability of the river Tisza is coupled with the main tributaries (the rivers Bodrog, Sajo, Bodva, Hernad, Szamos, Kraszna, Koros and Maros) and their high slope gradients that can cause rapid increase of the water level in the course of intense rainfall in the upper parts of the headwaters and the catchment areas (Fig. 8) (Nagy et al., 2010; Schweitzer, 2009).Fig. 9. 2.6. Multivariate environmental model and quantitative statistics In landscape archaeology, multivariate modelling is performed by integrating different GIS-based environmental and surface data sets (Groenhuijzen, 2019; Howey, 2011; Howey and Brouwer Burg, 2017; Kempf, 2019b, 2020; van Dinter, 2013). In most cases, the method is used to investigate movement corridors or to model accumulative and multi-layered environmental data surfaces that represent a classified permeability of spatial information (Kempf, 2019a; van Lanen et al., 2015a, 2015b). The multivariate landscape model identifies the areas that best fulfill the empirically addressed environmental parameters. In contrast to the validation of a deductively developed theory of human activity patterns in the landscape, an inductive approach is based on the conclusions of the case studies underlying site-specific research (Giiimil-Fariha and Parcero-Oubiha, 2015; Verhagen, 2019; Weaverdyck, 2019). For this purpose, large-scale surface models are tested against each other on different local scales to identify changes in the immediate landscape elements at continuous distances from the sites. This results in a pattern for different radii around each site, which allows determining the range at which significant changes in the composition of the landscape elements can be observed. This is the threshold value that indicates the transition between the environmental dependence of the site-catchment and the randomness of the periphery. 2.6.1. Site location analysis Site preference analysis and the analysis of surface conditions within a given catchment of a site are sound methods in archaeological research for some time now (Roper, 1979; Ullah, 2011; Vita-Finzi et al., 1970). The most strongly discussed issue, however, remains the determination M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 46.5 16.5 18.0 19.5 21.0 22.5 Excess Water Hazard in Hungary + + savo / v nstfft*. tíha;;' ////V, j viso MECE KOKI -^i .... A a* popiPS. //'/'yy n 4roreL ebvo PULE ■■• V, ;//■'''. /' / csat /. ' bud'/v //, VAX> A- '//f'WM ■ * 'V' "' %^K.9KEA mort ale; ,. bamA\Jbal. „__uvegi . lgcs^ , SZEH^AVAM6-116 hogo .A hoko , děor' + SEKU BORIv 24.0 _I_ A Neolithic site » Sand ZZl Fluvial system Excess water hazard _7J 0 7^ 20 ZJ 5 I I 25 I 10 ■ 30 15 no data 50 100 km (Michael Kempf, 2019; raw data provided by László Pásztor) Fig. 8. Inland excess water hazard map (IEW) from Hungary. The map shows frequently inundated areas with high flooding potential (>15), medium flooding potential (10 - 15) and low flooding potential (<10). The sand-covered regions in Hungary are characterized by low flooding vulnerability and dry conditions. The IEW data was kindly provided by Laszlo Pasztor (Bozan et al., 2018; Pasztor et al., 2015a, 2015b) and sand-covered areas were processed from (Pasztor et al., 2018). of the (fixed) distance (buffer) around a site, which can be spatial (distance) or spatio-temporal (walking distance). In this case, a different set of fixed distance buffers from 1 to 5 km were applied to estimate spatial components of the geological, pedological, and hydrologic components. Finally, a spatio-temporal consideration based on slope gradient, accessibility, and permeability has been applied within a fixed 3 km buffer around the sites, which represents the threshold, at which the surface conditions approximate random conditions. Consequently, a surface model to identify suitable site-ranges was created around the sites. The spatial analyses were performed using QGIS 3.6.0 (Open Source Geospatial Foundation Project, http://qgis.osgeo. )rg) and GRASS 7.6.0 software (Geographic Resources Analysis Support System, http://grass.osgeo.org). All statistical calculations were performed using R software (The R Project for Statistical Computing, www.r-project.org). All 49 sites were first analyzed for their spatial relationship to the soil units in Hungary (Pasztor et al., 2018). The chi-square goodness-of-fit test was applied to examine the frequencies of site distribution and soil dispersal (Kvamme, 2020). The null hypothesis is formulated as follows: Neolithic sites in Hungary were randomly distributed. The statistical evaluation can be achieved using this equation: * (0|. _Eif Where Of are observed and £( expected sites within the respective soil unit. The soil units are dispersed as follows: Chernozem (18,5%), brown forest soil (22,14%), meadow soil (26,5%), sand soil (11,25%), alluvial soil (5,2%), skeletal soil (3,2%), salt affected soil (6,34%), lithomorphic soil (4,56%), and peat soil (1,59%). The results show no significant deviation of the observed from the expected site distribution (X-squared = 13.884, df= 8,p-value = 0.08485). This evaluation, however, is point-based and does not allow to draw conclusions about the individual soil composition at variable distances around the site locations. To understand the preferences and particularly the soil components around the sites, which were modelled from GIS-surface calculations, a statistical analysis was carried out using a combination of GRASS and R software. First, the soil data was spatially analyzed using the r.neighbors function in GRASS. This function looks at each soil-cell of the raster input layer and considers the adjacent values assigned to that cell in a custom neighborhood. In this case, the assigned cells were located in a circular pattern around each cell. The function outputs a new raster layer where every cell is assigned a value, which is a function of the values of the surrounding cells. To calculate fixed radii with a cell size of 100 m, neighborhoods of 21, 41, 61, and 101 cells were chosen, which equals circular buffers of 1, 2, 3, and 5 km around each cell (the center cell is included in the calculation). In a second step, all Neolithic sites were assigned the soil value of their location using the point sampling tool plugin in QGIS. This outputs a csv-file with multiple point locations from a set of raster layers. Finally, a comparison dataset with 1000 randomly distributed points was generated, and every point was assigned the value of the respective soil raster layer based on the function described above. This comparison dataset was used to evaluate the expected and observed distribution of soil composition in different radii around the 49 Neolithic sites. For this reason, the two-sample Kolmogorov-Smirnov test (KS-test) was applied to compare the samples (Nakoinz and Knitter, 2016). This test works under the 0-hypothesis, that both samples are drawn from the same distribution. The test allows to check whether two samples derive from the same distribution (or population) or if they are different. However, it does not allow to draw conclusions about how or what these differences are. P-values range from 0 to 1 with strong significant differences in the distribution at 0.01 and no differences at 1. To enhance the statistical information of the two-sample KS-test, the Vargha-Delaney A test (VD A) was applied (Vargha and Delaney, 2000). Just like in the KS-test, the Neolithic sites were compared to the comparison data in a two-sample test. The R-package effsize (Efficient Effect Size Computation) created by Marco Torchiano includes the algorithm for the VD A test. Eventually, KS-test and VD A test were performed for both site datasets on the nine soil classes and over the chronological periods Early, Middle, and Late Neolithic as well as the Chalcolithic period, 9 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 \ i D \ Olu»l-JS<0)^OQO[li-IOziLUc5oH^P^^ <<»°SiuiOomuH«ini<^ -L- x LL L±J \> LL ^ S P 9 CO CO LU ( o r: o : PIS! a. lu Q z LU O □ O =J > ^ LU I P < LU N 0- CO CO CO ' O < O LU Q X CO b o p 10 I I] L n l" D i 1 -1 lu cd j 5< —i < < zi 5 O o m co co 0-l-0c'OO"jy30rLUO^LL U O Q LU LL Li)X — i o o Z(0IDUlOHiF:illODOIOO -JS2555 ''-o.cocococo'^l-1 5 tH O m O I andesite H granite H impure dolomite H impure limestone I I limestone I I mudstone I I organic rich sediment I sandstone HB tuff-breccia I I diamicton I I sand I I silt I I slate I I no data [water] 0,1 -0,15 -0,2-1— 1 km 2 km 3 km 5 km Fig. 9. Geological conditions in the site catchments of the 49 Neolithic sites in Hungary. Three buffer distances (r = 1 km, r = 2 km, r = 3 km) were selected and the respective areas covered by each geological unit were plotted for each site. The buffer of 1 km shows spatial significance of fluvial deposits (diamicton) and silty material (loess). With increasing buffer, the geological diversity increases. The VD A p-values confirm the preference for diamicton and silt, while sand is completely avoided, and the other geological units were not considered in the site location decision-making during the Neolithic period. 10 M. Kempf which allowed to track site location preferences as a function of soil distribution in variable distances around the sites (Table 4). The p-values range from 0 to 1. The plotted p-values of the VD A calculations were subtracted by 0.5 to generate significant values in the positive and negative value range (Fig. 12). Positive values indicate significant site preferences, values around 0 point toward no particular preference, and negative values indicate significant avoidance of the given soil distribution within the spatial range. The geological dataset from the EGDI database was processed using the same parameters. 2.6.2. Accumulative surface model Floodplain dynamics were simulated from buffered large rivers (river Danube and Tisza and the major tributaries) and the small scale hydrological system in Hungary with a fixed distance buffer of 20 m. Accordingly, the observed floodplain extent of the Corine riparian delineations was modelled for the major river courses (https://land. Copernicus.eu/local/riparian-zones; last accessed 09.12.2019). The results were merged to simulate a potential flooding vulnerability in Hungary based on soil composition, hydrological ecosystem, and streamflow dynamics (CLC, CRD 2019). Palaeochannels along the river Danube and the river Tisza were identified from satellite imagery, historical maps, and partly from the geological and pedological signatures. For this reason, Sentinel-2 multispectral images with a medium-resolution of 10 m in the visible range have been spectrally modified to obtain information about modern crop diversity and land-use patterns in the floodplains. Through the evaluation of vegetation indices (NDVI = Near Infrared - Red/Near Infrared + Red) (Lasaponara and Masini, 2006; Montandon and Small, 2008; Verhulst et al., 2009) that mirror the physical plant condition, patterns of modern cropland, sealed areas, forest cover, and wetlands were identified. In the satellite imagery, former palaeochannels and oxbow lakes are visible in the land-use patterns and the type and physical conditions of the vegetation cover, which mostly consists of forest or grassland coverage. Terrain roughness in Hungary was processed from a digital elevation model (DEM) with a 25 m spatial resolution that was downloaded free of charge from the United States Geological Survey (USGS). All slopes greater than 10° were considered unfavorable for agriculture due to their increased erosion vulnerability during heavy rainfall and weakly developed soil coverage. Furthermore, potential accessibility and permeability as well as crop irrigation possibility decrease with growing distance and slope gradient. That is a generalized limitation as the heterogeneous soil mosaic in Hungary shows very different erosion susceptibility. Soils that developed over loess deposits with a higher water storage capacity show different erosion behavior than shallow soils on lithic debris with little vegetation coverage. However, as has been pointed out by Siimegi et al. (2012), the various micro-regional and local floral refuges have produced a highly diverse soil development in the Holocene and a slope generalization to 10° allows for standardized digital analyses. From these components, an accumulative permeability and usability surface was calculated using the r.cost-function in GRASS. The surface represents the potential accessibility of the site catchments of each site. Permanent waterbodies, palaeochannels, the observed and the potential floodplain as well as high terrain roughness were accumulated to represent areas of high and low accessibility. 3. Results Variable distances around 49 Neolithic sites were tested for their surface conditions, soil texture and compositions as well as geological units, permeability, and accessibility. In this context, particularly geological conditions and soil composition play a major role in site location parameters in the immediate surrounding of the sites. Water availability further increases the potential land-use and settlement site opportunities. Journal of Archaeological Science: Reports 34 (2020) 102637 3.1. Geological and pedological site locations Geological conditions were measured in 1, 2, 3, and 5 km distance to the sites following the parameters described above. Table 3 shows the VD A statistics for the entire Neolithic. There is a clear differentiation between Holocene sand soils, which were significantly avoided by Neolithic groups and floodplain-related soil units. Most geological units did not play a significantly important role in the choice of potential settlement locations. Only Pleistocene loess-covered area, partly calcareous patches characterized by limestone or dolomite, and dia-micton showed significant preferences in the database. However, due to the rather coarse resolution and harmonization of the geological data, the preferences for alluvial sediments and loess-covered areas can only indicate an overall trend in potential land-use patterns. Consequently, the major soil type classification of Hungary based on the data provided by Pasztor et al. 2018 was spatially analyzed for site catchment compositions (Fig. 10, Fig. 11). Nine major soil types - according to climatic, geographical, and genetic basis - can be distinguished that show local subdivisions as reported from Micheli et al. (2006). The soil types consist of skeletal soils, shallow lithomorphic soils, sand soils, brown forest soils, Chernozem, peat soils, meadow soils, alluvial soils, and salt affected soils (Micheli et al., 2006; Pasztor et al., 2018). Hungary is mostly covered by brown forest soils, meadow soils and the various derivatives of Chernozem (Fig. 10). The Neolithic site distribution correlates spatially with the soil proportions, however there is a significant increase in site density on meadow soils (n = 23) and a rather unproportional distribution on Chernozems and brown forest soils. Skeletal, salt affected, sandy, peaty, and lithomorphic soils show the expected low site concentrations. The results were further evaluated for site-distance analysis of the nearest soil patches around the sites. Sandy soils, peat soils, lithomorphic soils as well as skeletal, and salt affected soils are dispersed in greater distance to the sites. Chernozem, meadow soils, and alluvial soils are whether direct location parameters of the sites or available in very close distance. There is significant support of distance relationships between sites and soil distribution within a radius of 3 km around the Neolithic sites in Hungary. The catchment analysis of the major soil types was processed for 1 km, 2 km, and 3 km (Fig. 11) distances for spatial visualizations and for 1, 2, 3, and 5 km for statistical analysis compared to a random point distribution (Fig. 12). There is considerable difference of soil composition between the ranges. The 1 km radius shows the dominance of meadow soils, Chernozems, and forest brown soils in the immediate surroundings of the sites. The soil mosaic increases in variability at a 2 km radius with continuous dominance of meadow soil patches. Unfavorable soil units like sand soils decrease in the ratio while a slight increase of salt affected soils can be detected in the assemblages. The 3 km radius intensifies the number of salt-covered soil variations with increasing variety of Chernozems and meadow soils in the near distance to each site. From these results, a site-catchment radius of maximum 3 km distance around each site was considered to best fulfill the potential activity ranges of Neolithic groups in the study area. 3.2. Soil composition at variable distances To evaluate the results from the surface model, statistical analyses using the KS test and the VD A statistics have been carried out. Despite the natural distribution proportions of the nine major soil types in Hungary and the predominance of meadow, forest and Chernozem soils, a preference for meadow, alluvial, and Chernozem soils in the Neolithic is significant. (Fig. 14). The p-values of both the K-S test and the VD A statistics show very significant results for the three soil units in a 1 km circular buffer around all Neolithic sites. Sand soils and surprisingly Forest Brown soils were clearly avoided at this distance, which strengthens the argument for statistical analysis and comparison data-sets. Skeletal, lithomorphic, peaty, and salt affected soils have no influence on the decision of a settlement site at 1 km distance 11 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Table 3 VD A p-values and KS test results for the geological units within 1, 2, 3, and 5 km distance around the Neolithic sites. Total Total Total Total Total Total Total Total Total Total Total Total 1 km 1 km 1 km 2 km 2 km 2 km 3 km 3 km 3 km 5 km 5 km 5 km Geological D KS p- VDA p- D KSp- VDA p- D KS p- VDAp- D KS p- VDAp- unit value value value value value value value value Volcanic 0.035247 1.0 0.5065612 0.039275 1.0 0.5047755 0.049345 0.9999 0.5012143 0.056785 0.9982 0.5148571 Clastic 0.0070493 1.0 0.4930 0.013092 1.0 0.49 0.016113 1.0 0.4885 0.021148 1.0 0.4860 Diamicton 0.17229 125 0.5707347 0.15215 0.2299 0.5657143 0.14095 0.3116 0.5628367 0.18149 0.09224 0.5630408 Calcareous 0.029266 1.0 0.5101939 0.059005 0.9969 0.5222143 0.087737 0.8649 0.5356735 0.092032 0.8239 0.5339796 Mudstone 0.055121 0.9989 0.4774694 0.087346 0.8684 0.4830306 0.075529 0.9527 0.4967245 0.084325 0.8941 0.4902041 Metamorphic 0.01638 1.0 0.5006735 0.015373 1.0 0.4986939 0.015373 1.0 0.4967347 0.017387 1.0 0.4894286 sediments Organic rich 0.032226 1.0 0.4958163 0.028197 1.0 0.4880918 0.034918 1.0 0.4808776 0.052038 0.9996 0.4781327 sediments Holocene sand 0.18402 0.08462 0.4185714 0.21169 0.03045 0.3944082 0.20994 0.03261 0.3872959 0.22317 0.01909 0.3782347 Pleistocene silt 0.1803 0.09603 0.5479184 0.16446 0.1599 0.5374796 0.11034 0.6204 0.5350918 0.11678 0.5474 0.5484694 measurement. This trend can be observed for all distances, although there is a significant increase in salt affected soils with greater distance from the site. Skeletal soils also increase, which is probably linked to increased distance from fresh water. The strong preference for meadow and alluvial soils further indicates that settlement and land-use strategies are strongly connected to the lower parts of the floodplains and the fertile and more humid soil compositions. Despite an increased flooding vulnerability, intensive agricultural exploitation and pasture can be assumed in close distance to the settlements. The chronological differentiation reveals a somewhat heterogeneous result. In the Early Neolithic period, meadow soils played a significantly major role in close distance to the settlement. At small-scale, Chernozem soils are important location factors as well while sand soils, lithomorphic soils and Brown Forest soils were avoided. Surprisingly, salt-affected soils were significantly located in a one to five kilometer distance around the sites, which could point towards salt exploitation or environmental change and extensively flooded and shallow lake areas with salt accumulation from weathering processes during the Early Neolithic. The Middle Neolithic provided the most homogeneous results, probably linked to the largest sample size. During this period, floodplain-related soils like alluvial and meadow soils are significantly distributed within all distances. Salt affected soils gain in importance with larger distances from the sites. All other soils were avoided or did not play a major role in the decisionmaking process of a settlement location during the Middle Neolithic. In the subsequent Late Neolithic, alluvial soils were again the predominant soil type in the catchments of the sites. Surprisingly, Chernozem soils cannot be considered significant location parameters according to the p-values. Sand soils were clearly avoided while lithomorphic and more stony soils can be found in variable significance around the sites. 3.3. Multivariate surface analysis The very close distance between the sites and the next available meadow and alluvial soil patches underlines the importance of rather wet or hydromorphic soils close to fresh water outside the periodically flooded areas. The local agricultural interaction radius was set to a 3 km buffer what supports the theory and the results of local cereal growing supported by pollen data (Bogaard et al., 2007). However, some sites are not situated on suitable soils. The sites at ABO, FAGA, DEOR, SAVO, SZEG, TIDO and TISO are located in modern urban agglomerations, which does not allow for the determination of a clear soil signal due to the modern bias. Only the site at PULE neither shows a modern urban bias nor in situ high quality soil conditions. However, extensive meadow soils and Chernozem are available in very close distance to PULE. The availability of running fresh water and the protection from flooding becomes visible in the spatial behavior of the sites. No Neolithic sites are located in the observed floodplain and the accumulated potential water surface. The intense agricultural utilization of the modern Hungarian landscape can be estimated from modern satellite imagery analyses. Most of the low-lying parts of the Great Hungarian Plain are used for (irrigated) crop cultivation, which caused dramatic shifts in the surface composition due to drainage or irrigation channel construction. Furthermore, the modern river system has experienced massive regulations and channeling activity during the 20th century what makes it difficult to draw conclusions about premodern environmental dynamics from modern datasets. However, the strong modern overprint that is visible in the 3 km buffer around the archaeological sites would suggest a hypothetical site-continuity in agriculturally favorable areas with high quality soil properties and low environmental vulnerability. On the other hand, the archaeological record used in this paper could also be significantly biased by the modern built-up change, which created an Neolithic distribution in areas that are spatially congruent with recent earth movements, extensive surface re-modelling, and an active archaeological survey activity. Both results can be considered a realistic contribution to potential Neolithic site dispersal in Hungary. It is however noteworthy, that the modern urban agglomerations are often spatially identical with the archaeological record. Terrain roughness and landscape permeability were assessed via a digital elevation model and the calculated slope gradients >10°. Most of the sites are located in flat areas with no steep slopes. Some sites do show considerable terrain roughness in the catchment areas. This is mostly due to their location outside the floodplain and on palaeolevees along the valleys. A landscape accessibility surface was processed from the hydrolog-ical and topographical landscape analyses. Fig. 13 shows an example of the accumulative surface with a fuzzy gradient from high to low. A high value is equal to high landscape accessibility and enables an individual to travel from one point to another without expending a high amount of energy. Due to the fact, that these surfaces include hydrological information and the spatial extent of the observed/accumulated flooding zone, the limitations of the movement corridor within the catchments varies greatly depending on the location close to a river or the shoreline of Lake Balaton. Most sites are located within a reasonable movement corridor that allows for high mobility at low-energy expenditure. Several sites show strict limitations due to the accumulation of the hydrological situation in the vicinity (e.g. TISO and BOVO) or the adjacent topographical conditions (e.g. BUD and the BAM, BAL, ALE and MORT complex). It becomes clear that the terrain roughness creates movement corridors along the valleys which fit the pedological conditions and the accumulated alluvial deposits that produce rather fertile soils in comparison to the lithic or sandy soils of the adjacent hills (e.g. SEKU, BORK). The site location parameters of rapid fresh-water access and fertile soils in a highly permeable landscape determine the most suitable natural prerequisites. Finally, a potential model for land-use opportunities around the 49 Neolithic sites in Hungary was processed from the accumulative site-catchment analyses, the accumulative potential water surfaces, 12 Table 4 Table showing the results of the Kolmogorov-Smirnov test (D, p-value) and the Vargha-Delaney A test (p-value) for nine soil classes in Hungary. The soil values were created using the r.neighbors function in GRASS, which makes each cell category value a function of the category values assigned to the cells around it. Neighborhoods of 1,2,3, and 5 km from each soil raster cell were chosen and the values for the chronological periods from Early, Middle, and Late Neolithic as well as the Chalcolithic were calculated. Soil class Total 1 km D Total 1 km KS p-value Total 1 km VDA p-value Total 2 km D Total 2 km KS p-value Total 2 km VDA p-value Total 3 km D Total 3 km KS p-value Total 3 km VDA p-value Total 5 km D Total 5 km KS p-value Total 5 km VDA p-value Chernozem Forest Brown soil Meadow soil Sand soil Peat soil Alluvial soil Lithomorphic soil Skeletal soil Salt-affected soil 0.19643 0.14337 0.05439 0.2923 0.5655204 0.4483469 0.2191 0.14337 0.02256 0.2923 0.5843673 0.4575714 0.24729 0.13237 0.006607 0.3863 0.5818367 0.4641735 0.21988 0.14159 0.02185 0.3062 0.5805408 0.4670306 0.27261 0.10114 0.014633 0.25416 0.071 0.001931 0.7257 1.0 0.004787 0.9726 0.6319184 0.4513061 0.5037347 0.6119898 0.5036633 0.23884 0.15051 0.07349 0.2498 0.046 0.009697 0.2405 0.9625 0.00588 1.0 0.6173367 0.4196122 0.5344694 0.6107041 0.5011122 0.23865 0.17792 0.063714 0.23543 0.043898 0.009777 0.1039 0.9914 0.01128 1.0 0.6068469 0.4178776 0.5297755 0.6147041 0.5094796 0.21592 0.17651 0.065449 0.22861 0.086347 0.02567 0.1089 0.9882 0.01515 0.8771 0.6045102 0.4218367 0.5199592 0.6249388 0.5288571 0.090898 0.055 0.835 0.9989 0.518602 0.5043673 0.11808 0.098571 0.5327 0.7545 0.5268163 0.5267653 0.12512 0.16261 0.4575 0.1690 0.5502959 0.5602347 0.19553 0.21888 0.05621 0.02277 0.5686122 0.5861327 Soil class Early Neolithic 1 km D Early Neolithic 1 km KS p-value Early Neolithic 1 km VDA p-value Early Neolithic 2 km D Early Neolithic 2 km KS p-value Early Neolithic 2 km VDA p-value Early Neolithic 3 km D Early Neolithic 3 km KS p-value Early Neolithic 3 km VDA p-value Early Neolithic 5 km D Early Neolithic 5 km KS p-value Early Neolithic 5 km VDA p-value Chernozem Forest Brown soil Meadow soil Sand soil Peat soil Alluvial soil Lithomorphic soil Skeletal soil Salt-affected soil 0.40733 0.3910 0.2757 0.3215 0.6486667 0.35275 0.39233 0.3820 0.3176 0.3489 0.62675 0.3848333 0.39533 0.3820 0.3089 0.3489 0.6231667 0.4181667 0.37267 0.3450 0.3789 0.4767 0.64275 0.4165 0.55633 0.2540 0.7400 0.19433 0.1260 0.04985 0.8363 1.0 0.9779 1.0 0.77575 0.3730 0.4630 0.5715 0.4370 0.5420 0.3750 0.21133 0.11033 0.17890 0.06014 0.3713 0.9527 1.0 0.9910 0.7916667 0.3125 0.5785 0.5310 0.4105 596 0.3980 0.15833 0.1690 0.1990 0.0289 0.3013 0.9983 0.9957 0.9722 0.7856667 0.3206667 0.4639167 0.56075 0.4589167 0.5430 0.4140 0.2620 0.35433 0.2510 0.05937 0.2583 0.8076 0.4423 0.8466 0.7428333 0.3356667 0.5154167 0.61225 0.5604167 0.2180 0.19933 0.9394 0.9718 0.3910 0.5623333 0.2810 0.3260 0.7340 0.5504 0.5775833 0.6033333 0.3330 0.42933 0.5228 0.2216 0.5870 0.7040 0.4130 0.5250 0.2609 0.07467 0.56675 0.7438333 Soil class Middle Neolithic 1 km D Middle Neolithic 1 km KS p-value Middle Neolithic 1 km VDA p-value Middle Neolithic 2 km D Middle Neolithic 2 km KS p-value Middle Neolithic 2 km VDA p-value Middle Neolithic 3 km D Middle Neolithic 3 km KS p-value Middle Neolithic 3 km VDA p-value Middle Neolithic 5 km D Middle Neolithic 5 km KS p-value Middle Neolithic 5 km VDA p-value Chernozem Forest Brown soil Meadow soil Sand soil Peat soil Alluvial soil Lithomorphic soil Skeletal soil Salt-affected soil 0.25047 0.14435 0.0323 0.4997 0.5906324 0.4520294 0.30188 0.14459 0.004991 0.4976 0.6238382 0.4430 0.33794 0.16118 0.001095 0.3602 0.63125 0.4465294 0.28053 0.15218 0.01131 0.4316 0.6238676 0.4518824 0.27141 0.079176 0.040588, 0.21935 0.0740 0.01574 0.9861 1.0 0.08448 0.9938 0.6306176 0.4861765 0.5055882 0.5964853 0.4797794 0.23165 0.11859 0.058471 0.23059 0.069765 0.05867 0.7442 0.9999 0.06059 0.9972 0.6079853 0.4661324 0.5129559 0.6013529 0.4812059 0.20365 0.11459 0.060647 0.25224 0.084176 0.1307 0.7810 0.9997 0.03047 0.9739 0.5926471 0.4675735 0.5068824 0.6211176 0.4767647 0.21412 0.12 0.091882 0.27271 0.069176 0.09808 0.7309 0.9441 0.01503 0.9975 0.5838529 0.4713676 0.4731471 0.6535294 0.4929118 0.061471 0.082118 0.9997 0.9796 0.4973971 0.5271912 0.066882 0.19941 0.9985 0.1463 0.4764412 0.5710882 0.078471 0.22024 0.9874 0.08236 0.5020 0.5839559 0.11329 0.23629 0.7925 0.05085 0.5218824 0.6050 Late Neolithic 1 km Late Neolithic 1 km Late Neolithic 1 km Late Neolithic 2 km Late Neolithic 2 km Late Neolithic 2 km Late Neolithic 3 km Late Neolithic 3 km Late Neolithic 3 km Late Neolithic 5 km Late Neolithic 5 km Late Neolithic 5 km (continued on next page) Table 4 (continued) Soil class Total 1 km D Total 1 km KS p-value Total 1 km VDA p-value Total 2 km D Total 2 km KS p-value Total 2 km VDA p-value Total 3 km D Total 3 km KS p-value Total 3 km VDA p-value Total 5 km D Total 5 km KS p-value Total 5 km VDA p-value Soil class D KS p-value VDA p-value D KS p-value VDA p-value D KS p-value VDA p-value D KS p-value VDA p-value Chernozem Forest Brown soil Meadow soil Sand soil Peat soil Alluvial soil Lithomorphic soil Skeletal soil Salt-affected soil 0.24 0.15 0.6185 0.9791 0.48655 0.51755 0.2120 0.1270 0.7651 0.9972 0.4888 0.51805 0.2 0.1410 0.8233 0.9893 0.50045 0.5142 0.2440 0.1770 0.5974 0.9157 0.5051 0.4861 0.1770 0.1540 0.0930 0.3780 0.2930 0.9157 0.9730 1.0 0.1181 0.3632 0.5438 0.4206 0.5160 0.66285 0.62595 0.3370 0.2750 0.09 0.3930 0.3010 0.2108 0.4424 1.0 0.09392 0.3310 0.5596 0.3620 0.52765 0.66465 0.61485 0.2690 0.3230 0.0820 0.32 0.3160 0.4708 0.2529 1.0 0.2627 0.2762 0.55385 0.34585 0.49135 0.63745 0.6402 0.2520 0.31 0.2260 0.3180 0.3010 0.5557 0.2973 0.6926 0.2693 0.3310 0.5710 0.3406 0.56995 0.60525 0.6172 0.2940 0.13 0.3590 0.9962 0.6052 0.4796 0.2820 0.1810 0.4105 0.9019 0.5739 0.42385 0.29 0.17 0.3757 0.9371 0.59285 0.4718 0.3680 0.1820 0.1369 0.8983 0.61625 0.52075 Soil class Chalcolithic 1 km D Chalcolithic 1 km KS p-value Chalcolithic 1 km VDA p-value Chalcolithic 2 km D Chalcolithic 2 km KS p-value Chalcolithic 2 km VDA p-value Chalcolithic 3 km D Chalcolithic 3 km KS p-value Chalcolithic 3 km VDA p-value Chalcolithic 5 km D Chalcolithic 5 km KS p-value Chalcolithic 5 km VDA p-value Chernozem Forest Brown soil Meadow soil Sand soil Peat soil Alluvial soil Lithomorphic soil Skeletal soil Salt-affected 0.23333 0.22433 0.9015 0.9249 0.4979167 0.4483333 0.17933 0.21533 0.9908 0.9449 0.5105 0.4244167 0.2280 0.21533 0.9158 0.9449 0.4836667 0.4579167 0.2830 0.17833 0.7259 0.9914 0.5021667 0.4811667 0.26033 0.13467 0.096667 0.21333 0.088667 0.8137 0.9999 1.0 0.9489 1.0 0.5183333 0.4721667 0.5408333 0.57125 0.5179167 0.25533 0.20833 0.3640 0.2720 0.11567 0.8316 0.9580 0.4082 0.7696 1.0 0.5424167 0.4204167 0.6533333 0.6059167 0.50025 0.25133 0.18933 0.3250 0.1790 0.10667 0.8455 0.9831 0.5544 0.9910 1.0 0.5499167 0.4390 0.6281667 0.5714167 0.4820833 0.23933 0.23133 0.2170 0.1510 0.13767 0.8841 0.9070 0.9415 0.9992 0.9999 0.5461667 0.4604167 0.58125 0.49175 0.5250 0.1320 0.2190 0.9999 0.9372 0.4703333 0.3905 0.15933 0.2030 0.9981 0.9666 0.4926667 0.46625 0.14267 0.1290 0.9997 1.0 0.4575833 0.5123333 0.1530 0.1510 0.9999 0.9992 0.51975 0.49175 soil M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Brown forest soils Chernozem Salt-affected soils 25k-^20k- Lithomorphic soils Sand soils Skeletal soils Alluvial soils Peat soils Meadow soils ■ 15k- 5k- (a) 20592 W 24645 Q 17209 j 10459 \ 4832 03643 ©4243 5895 1479^^ t-1-1-1-1-1-1-nr-r # # # # / # # ß V/////£ K ii ^ ^ t— ^ ^ •> 0) ^ O Q (5 - . _ „ ^ _ . __x _ _ _ ua:>DijQCD[u^?D:iii(5^(!i^oOOju^aiii SOODQogwmm<"-<<^LULL]caDXC0t-lCL yp 10 high > 30 Accessibility High 0 1 2 km Fig 13. Exemplary visualization of the landscape analyses and finally land-use opportunities at Tiszalok-Hajnalos (TIHA), located at the river Tisza. Displayed are soil patches within a 3 km buffer around the site, accumulated floodplain dynamics and palaeochannels, elevation and slope gradients, and multivariate accumulative surface based on an accumulative flooding area, the hydrological system, and terrain roughness (slopes > 10°). also on the immediate topography and thus the ultimate landscape permeability around the site. Although the supra-regional understanding of ecological feedback and interrelationships is inevitable for the interpretation of large natural environmental networks, an additional site analysis on the small-scale creates knowledge of the complex relationships between geological strata, soil coverage, surface development, and the hydrological system that control local human behavior within a local community. Depending on the extent of the research area, inaccurate terminology of scale and the unprecise implementation of site catchments can rapidly lead to misinterpretations of human-environment and human-human interactions (Harris, 2017). No-data corridors were created that either do not contain archaeological records or are considered lacking the potential of pre-modern development. Both considerations are misleading. With reference to spatial modelling, the scale-dependent critical analysis of topographic data is a way to getting closer to understanding potential pre-modern environmental conditions. 4.1. Holocene landscape dynamics Holocene landscape and surface dynamics are ultimately linked to the hydrological system, soil development, climate oscillations, and vegetation shifts (Magyari et al., 2010). It has frequently been argued that the Hungarian hydrological system experienced major relocation processes in the Late Pleistocene and the Early Holocene. The river Tisza and the river Danube both changed their river-course several kilometers 18 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 ABO MORT. KOKI Fig. 14. Potential land-use around Neolithic sites in Hungary based on accumulative potential water surface, landscape permeability, and soil units. to the west - what created sandy alluvial patches (e.g. in the DTI) that were locally covered with Late Pleistocene/Early Holocene loess on uneroded levees and fine-grained clayey material with high organic carbon content in the low-lying parts of the former palaeochannels. The Early Neolithic Koros culture and the Cri§ culture that occupied broad areas along the middle and the upper part of the river Tisza, however, can be traced on both sites of the river-course (Domboroczki and Raczky, 2010; Molnar and Siimegi, 2007), which points towards a stable riverbed system after the Early Neolithic. The river Tisza is bordered to the west by the subsequent alluvial deposits of the elevated DTI. The Interfluve does not show significant archaeological records underlying this study. However, it has been argued by Banffy (2012) that this is not necessarily connected to the limited land-use potential and the increasing aridity of the DTI but could also be due to limited archaeological research in the area (Banffy, 2012). Consequently, the poor archaeological record of the DTI is not mandatorily attached to erosion and accumulation processes of the west-moving river Tisza floodplain. Magyari et al. (2010) argued that persistent steppe vegetation with temperate deciduous wooded steppe grasslands and saline tall-grass 19 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 MEKO TIBO Fig. 14. (continued). meadows developed in the Great Hungarian Plain throughout the Ho-locene. The steppe-vegetation paradigm, as labelled by Siimegi et al. 2019, however, can probably not be confirmed for entire Hungary and pollen analyses alone may not yield all the required information for a comprehensive palaeovegetation reconstruction (Siimegi et al., 2013b). Because Hungary was not glaciated during the last glacial period, cold-resistant species survived in micro-refuges in the southern part of the Carpathian Basin where continuous soil development took place (Siimegi et al., 2019, 2013b; Siimegi, Persaits and Gulyas, 2012). At the transition to the Holocene, a mosaic of boreal-type forest-steppe vegetation spread to the north, which was highly vulnerable to natural fires further accelerating the development of a forest-steppe coverage (Siimegi et al., 2012). In the Mid-Holocene, climate reconstructions propose increased warming and pollen analyses revealed a dramatic shift from coniferous to deciduous forest in the Early Holocene (Kasse et al., 2010). However, as Feurdean (2005) pointed out for the northwestern part of Romania (Gutaj Mountains), Corylus became dominant in the forest while all other deciduous tree species declined between 7300 and 3750 BCE, probably due to cooler and drier conditions and a reversal to cooler and moister conditions from about 4800 BCE (Feurdean, 2005). Pollen analyses from the southern Carpathians further demonstrated Early Holocene warming and forestation followed by rapid summer cooling and increased moisture availability in the Mid-Holocene (Magyari et al., 2009). Vegetation response may thus be 20 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 MECE FEB TOLM Neolithic site Soil units Skeletal soil Sand soil Lithomorphic soil Brown Forest soil Chernozem .....j 5 km radius Salt affected soil I I 3 km radius Meadow soil = Waterways Peat soil y/%fi« Acc. potential water surface Alluvial soil | | Palaeochannel Fig. 14. (continued). 3 km controlled by the topographical situation and cooler and drier conditions may have prevailed in the Great Hungarian Plain throughout the Early and Mid-Holocene while the Carpathian Mountains experienced increased moisture availability. According to Kiss et al. (2015), groundwater dropping in the DTI led to forest decline and replacement by steppe vegetation during the Boreal Phase (Kiss et al., 2015; Siimegi et al., 2012). The subsequent Atlantic Phase was characterized by reforestation under warmer and moister conditions. However, the second half of the Atlantic Phase shows evidence for drier conditions and reactivation of aeolian activities. River dynamics during that time show strong avulsion events that most likely were triggered by tectonic subsidence processes and not by environmental shifts (Kiss et al., 2015). The increased vegetation coverage in the early Atlantic Phase thus would have prevented soil erosion and superficial runoff in the catchment areas of the tributaries of the river Tisza, which would result in decreased sediment input and fewer flooding events. Intra-floodplain channel erosion would rather stabilize the smaller Holocene channels of the meandering river (Kasse et al., 2010). However, Neolithic and the subsequent anthropogenic landcover change could have had a significant impact on the sediment transport and the fluvial regimes of the Hungarian river system. Deforestation activity and slash and burn agriculture may have not only had a strong effect on the soil assemblages and the transformation of Chernozems on loess-covered ridges but could have also led to the reactivation of palaeochannels during cooler and drier conditions (Heyvaert et al., 2012; Heyvaert and Walstra, 2016; Kabala et al., 2019; Lorz and Saile, 2011; Stouthamer and Berendsen, 2000). 4.2. Neolithic soil development and land-use strategies in the Carpathian Basin Over the past decades, the importance of loess in the spread of Neolithic agriculture has been intensely discussed (Bakels, 2014; Gro-nenborn, 1999; Hedges et al., 2013; Kreuz, 2007; Saqalli et al., 2014; Sielmann, 1972). The spatial congruence of the loess belt and the development of early farming techniques is often used to explain different subsistence strategies, the theory of shifting cultivation, or the establishment of permanent and manured fields and settlement patterns from the Carpathian Basin to the northern margins of the river Rhine (Bogaard et al., 2013; Brounen et al., 2010; Gronenborn, 1999; Rivollat 21 M. Kempf et al., 2016; Saqalli et al., 2014; Vanmontfort et al., 2010). However, the soil formation on loess varies locally due to variations in climate, the nature of the soil parent material, wildfire, and the impact of human land-use (Catt, 2001; Feurdean et al., 2020; Schmidt et al., 1999). Chernozems in Hungary are supposed to develop under steppe or deciduous forest vegetation and in combination with the subsequent human deforestation and utilization activity. The extensive development of the different types of Chernozems in Hungary is more complicated and locally dates back to the Holocene (Michéli et al., 2006; Smalley et al., 2011; Sumegi et al., 2013b; Vysloužilová et al., 2015). The Hungarian major soil type classification system distinguishes nine different soil characteristics (Michéli et al., 2006; Pásztor et al., 2018). According to Michéli et al. (2006), the various Chernozem types are frequently intermixing with each other, depending on groundwater level, vegetation type, and soluble salts, which generates a broad variety of subgroups such as meadow Chernozems or Chernozem brown forest soils. Neolithic people, therefore, were not forced to prefer meadow soils in the close vicinity to Chernozem but rather focused on what was locally available in close distance to freshwater (Hedges et al., 2013). Even though the levees could have been broadly covered with loess, the generally sandy texture of the levee would rather not support agriculture due to the high surface water permeability, low water storage capacity, and the low-lying groundwater level (Sumegi et al., 2012). Crop cultivation on free-draining soils in considerable distance to a Neolithic settlement and under dry climatic conditions does not seem reasonable -as reported from other Neolithic sites (Araus et al., 1999; Roberts and Rosen, 2009; Wallace et al., 2015). Indications for crop planting in naturally wet soils that experienced periodical flooding and deposition of fine-grained marshy silt and relocated loess exists from the Near East (Araus et al., 1999). Yields on these fertile soils would have been higher than on the dry locations of the levees. Sumegi et al. (2013a), however, argue that in the alluvial Polgár island, a loess-covered lag surface in the upper Tisza region, Neolithic land-use took place on the Chernozem-covered palaeo-levees of the river (Sumegi et al., 2013c). The question remains, if this is due to the high flood vulnerability of the lower floodplain or the fertility of the Chernozem on top of the levees. According to the authors, the lower parts were still covered with Chernozem, hydromorphic Chernozem, and hydromorphic forest soils that developed under gallery forest vegetation (Sumegi et al., 2013c). Even though the mosaic soil patterns remain visible, the Neolithic land-use and deforestation had a considerable impact on the landscape, which is mirrored in the expansion of hydromorphic Chernozem soils and the disappearance of the forest soils (Sumegi et al., 2013c). In addition, salt affected soils played a certain role in site location preferences of Neolithic farmers in Hungary (Bánffy, 2013, 2015, 2019). Saline soils are naturally tied to poor water drainage capacity and periodical waterlogging conditions. Generally, that does not enable intensive agricultural utilization. The amount of soluble salts in the soil and an evapotranspiration that exceeds the average precipitation determine the development of the strongly saline Solonchaks or the ameliorated steppe-meadow Solonetz Soils (Michéli et al., 2006). In the rather dry climatic conditions of the DTI, and especially in the river Tisza floodplain, large-scale salt affected areas are visible in the soil distribution maps (Fig. 13 b). Most of the sites are located in considerable distance to the next non-cultivable salt affected plains. However, salt-exploitation in the Neolithic Carpathian Basin is an important feature in the Neolithisation process due to the shift from a salt- and protein-rich meat-based diet towards a mixture of cereals with a low salt-content (Bánffy, 2013). Furthermore, salt is a substantial component of livestock breeding (Harding, 2015). A potential use of the extensive saline soils in parts of the Carpathian Basin is thus reasonable. Despite the local Late Pleistocene/Early Holocene loess-coverage of the uneroded plateaus, the lower meadows would have been more suitable for agriculture. Soils that were dominated by a locally high aquifer or prone to periodical flooding would be classified as meadow Journal of Archaeological Science: Reports 34 (2020) 102637 soils in the typology - probably caused by the constant deposition of fine-grained clayey material and organic matter in topographical depressions (Michéli et al., 2006). Slightly elevated meadow soils with lower flooding vulnerability would consequently experience strong agricultural utilization, which would have strengthened the human-induced transformation of the parent-soil into a mature Chernozem (Barczi, Golyeva and Petó, 2009; Kleber et al., 2003; Strouhalová et al., 2019) . In this context, strongly fragmented charred organic carbon (COC) and black carbon (BC) deriving from burning vegetation or by wind-blown transportation can be incorporated into the soil texture, which alters the composition, color, humic acid content, and soil organic matter (Schmidt et al., 1999; Schmidt and Noack, 2000). The carbon content from natural wildfires can be significantly increased through grassland or forest burning management. Although the environmental contexts cannot entirely be compared to the Neolithic soil development in the Carpathian Basin, Skejmstad et al. (1997) report from up to 30% increase of COC from artificially and continuously burnt Australian soil surfaces (Skejmstad et al., 1997) and Schmidt et al. (1999) confirm up to 45% of bulk organic carbon in chernozemic soils from Germany. Particularly the results by Schmidt et al. (1999) point towards Chernozem development in relation to increased COC content and burning activity from about 8.000 BP onwards, which aligns with the early Neolithic landscape transformation - independently from large-scale climatic conditions and vegetation patterns (Schmidt et al., 1999). Furthermore, the anthropogenic overprint of the ecosystem and the Neolithic landscape development was additionally reinforced by the increased presence of BC in the A-horizon of the cleared areas, which decreased regional surface reflectance and the albedo through enhanced 'blackness' of the surface (Hammes et al., 2007). Because of its pyro-genic origin and carbonaceous substance, BC is highly resistant to chemical and thermal degradation and has a long and stable residence time after accumulation and sedimentation, which highlights the importance of BC, fire, and anthropogenic land-use in the development of Chernozem across Europe (Hammes et al., 2007; Kleber et al., 2003; Preston and Schmidt, 2006; Schmidt et al., 2002; Thiele-Bruhn et al., 2014) and in particular in the Carpathian Basin during the Atlantic stage. This is in accordance with recent discussion about potential vegetation transition and recovery and the impact of herbivores on the natural vegetation as well as the maintenance of broad open landscape patches through fire and human impact (Doppler et al., 2017; Feurdean et al., 2020; Gerling et al., 2017; Kirby, 2004; Kreuz, 2007; Ochs et al., 2020) . Chernozem is the result of a long maturation of the soil organic matter. The black carbon can be a part of the soil organic matter, but this is not unconditional. Chernozem is in large parts of the world a zonal soil, where its formation is connected to the steppe vegetation, intensive bioturbation, carbonate parent material, and many other factors, which are not fully understood. Generally, the black color is due to the high organic matter content and the black carbon originating from vegetation burning can contribute to the soil organic matter. Across Europe, however, there is lack of direct evidence and analyses are still missing. In addition to locally and regionally unclear soil development, the modern lowering of the groundwater level would further intensify soil degradation and erosion in most parts of the Great Hungarian Plain. Modern soil erosion during the past 100 years has locally removed the top-soil and the loess layers, which triggered the outcrop of the underlying geological basement (Kerényi, 1994) - making it particularly difficult to evaluate the original loess cover and the thickness of the prehistoric soil layers. Furthermore, a continuous Holocene soil development in the Carpathian Basin would implicate continuous climatic conditions and hence create homogeneous soil thickness with no interspersed palaeosoil layers or dust accumulations. Kiss et al. (2015) report from reactivated dust accumulation during drier periods in the Early Holocene and throughout the Atlantic Phase (Kiss et al., 2015). That fits the results from the Loess Plateau in China where Holocene dust deposits alternate frequently with weakly developed palaeosoil layers (Maher 22 M. Kempf et al., 2003). With regard to Mid-Holocene climate variability in the Carpathian Basin and the trend towards cooler, drier, and windier conditions during the late Atlantic Phase, forest-cover decline, and the reactivation of sandy and silty deposits would also have reworked aeolian dust transportation. Weakly developed Chernozems would have then been covered by repeated loess sedimentation. All Neolithic sites in this paper are situated outside the potential accumulated floodplain, which indicates protection from extensive flooding and loss of harvest and livestock. However, one can easily argue that our modern perception of the archaeological record is strongly biased by the findability of archaeological traces in modern land-use corridors (e.g. the massive built-up change) and in areas that did not experience strong surface cover modifications through erosion or accumulation processes. Hence, the fact that Neolithic records can be found outside the (modern and pre-industrial) erosion zone of the river Tisza and the river Danube seems logical. Moreover, as visible in modern satellite imagery, the whole landscape is ultimately restructured by recent anthropogenic overprints and interventions in the natural ecological balance. Drawing large-scale conclusions from modern datasets needs to consider the methodical limitations of the data. 5. Conclusions Assessing Neolithic agricultural strategies through multivariate environmental analyses has proven to be a useful tool to understanding the scale of past human behavior. The integration of comprehensive ecological and geographical links and feedbacks in landscape archaeological research enables not only an interpretation of the spatio-temporal extent of human activity ranges but also the evaluation of palaeosurface changes and landcover modifications triggered by human-environment interaction. This approach represents an important link in interdisciplinary research and provides new insights in soil development, vegetation history, and palaeohydrological system transformation. In this article, 49 Neolithic sites have been investigated in terms of site location parameters and particularly soil preferences. Site-catchment analyses revealed a strong spatial relationship between site selection and soil properties with a significant preference for rather humid meadow and alluvial soils in close distance to running fresh water. The theory of a strong interrelationship between early agricultural strategies and Chernozem soil on loess-covered areas, alluvial fans, and palaeolevees cannot be completely traced throughout the Neolithic Period. This article furthermore suggests that the Neolithic impact on the surface cover could have triggered or enhanced soil development and thus would have contributed to the modern extensive Chernozem coverage in the Carpathian Basin through intensified land-use, clearing, and carbon input. The Neolithic preferences for meadow and alluvial soils would have therefore been much stronger and the modern soil data would be biased by the strong agricultural overprint of the past 7000 years. Future pedological analyses and the continuous incorporation of modern scientific data sampling in Neolithic excavations will clarify the picture of soil development during human-environmental interaction in the Carpathian Basin. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements First, I would like to emphasize the constructive remarks of the two reviewers, which helped to strengthen the structure behind the article. Furthermore, I am very grateful to Eszter Banffy (RGK-DAI Frankfurt) and Kurt Alt (Basel/Krems) who initiated the DFG-project (grant number Al 287/10-1) and provided access to their site database. Margaux Journal of Archaeological Science: Reports 34 (2020) 102637 Depaermentier (Basel) added important comments to the catchment analysis and Barbora Strouhalová (Prague), Eva Rosenstock (Berlin), Pál Siimegi (Szeged), and Johannes Tintner (Vienna) strongly supported the discussion about Chernozem development and provided a lot of helpful references. Eli J. A. Weaverdyck helped to strengthen the statistical analysis. Finally, I would like to particularly thank László Pásztor (Budapest) and his team who provided access to invaluable soil data that enabled the analyses. Datasets Corine Landcover CLC: https://land.copernicus.eu/; last accessed 30th July 2019. Corine Riparian delineations: https://land.copernicus.eu/local/ riparian-zones; last accessed 09th December 2019. https://map.mbfsz.gov.hu/; last accessed, 23rd July 2019. http://map. georgikon. hu/en/mezogazdasagi-talajterkep; last accessed 17th June 2019. EGDI WFS layer: http://mapsrefdev.brgm.fr/wxs/lGE/EGDI_l-M_INSPIRE_geolUnits; last accessed 05th August 2019) References Acksel, A., et al., 2017. Human activity formed deep, dark topsoils around the Baltic Sea. Geoderma Region. 10, 93-101. https://doi.Org/10.1016/j.geodrs.2017.05.005. Acs, F., Breuer, H., Skarbit, N., 2015. Climate of Hungary in the twentieth century according to Feddema. Theor. Appl. Climatol. 119 (1-2), 161-169. https://doi.org/ 10.1007/S00704-014-1103-5. Araus, J.L. et al. (1999) 'Crop water availability in early agriculture: evidence from carbon isotope discrimination of seeds from a tenth millennium BP site on the Euphrates', Global change biology, 5, pp. 201-212 (Accessed: 10 December 2019). Bakels, C, 2014. The first farmers of the Northwest European Plain: some remarks on their crops, crop cultivation and impact on the environment. J. Archaeol. Sci. 51, 94-97. https://doi.Org/10.1016/j.jas.2012.08.046. Balla, D., Novák, T J., Zichar, M., 2016. Approximation of the WRB reference group with the reapplication of archive soil databases. Acta Universitatis Sapientiae, Agriculture and Environment 8 (1), 27-38. https://doi.org/10.1515/ausae-2016-0003. Bánffy, E. (2004) The 6th Millennium BC boundary in Western Transdanubia and its role in the Central European transition (The Szentgyórgyvólgy-Pityerdomb settlement). (Varia Arch. Hung., 15). Budapest. Bánffy, E., 2006. Eastern, Central and Western Hungary - variations of Neolithisation models. Documenta Praehistorica 33, 125-142. https://doi.org/10.4312/dp.33.13. Bánffy, E. (2012) 'South Western Kórós culture settlement in the Danube-Tisza interfluve: Szakmár-Kisiilés', in Anders, A. and Siklósi, Z. (eds.) The Kórós culture in eastern Hungary. (BAR International series, 2334). Oxford: Archaeopress, pp. 53-68. Bánffy, E. (2013) 'Tracing 6th-5th millennium BC salt exploitation in the Carpathian Basin', in Harding, A. and Kavruk, V. (eds.) Explorations in salt archaeology in the Carpathian Zone. (Archaeolingua, 28). Budapest: Archaeolingua Alapítvány, pp. 201-207. Bánffy, E., 2015. The beginnings of salt exploitation in the Carpathian basin (6th-5th millennium BC). Documenta Praehistorica 42 (15, pp). https://doi.org/10.4312/ dp.42.13. Bánffy, E. (2019) First Farmers of the Carpathian Basin: Changing patterns in subsistence, ritual and monumental figurines. Oxford, Philadelphia: Oxbow books (Accessed: 2 August 2019). Bánffy, E. and Bognár-Kutzián, I. (2007) The Late Neolithic Tell settlement at Polgár-Csószhalom, Hungary: The 1957 excavation. (Archaeolingua Central European series, 4). Oxford: Archaeopress. Bánffy, E., Juhász, I. and Siimegi, P. (2008) 'A prelude to the Neolithic in the Balaton region - new results to an old problem', in Spataro, M. and Biagi, P. (eds.) A short walk through the Balkans: the first farmers of the Carpathian Basin and adjacent regions: Proceedings of the Conference held at the Institute of Archaeology (UCL) on June 20-22nd, 2005. Atti della Societa per la Preistoria e Protostoria della Regione Friuli-Venezia Giulia, XVI. Trieste, pp. 223-237. Barczi, A., et al., 2006. Survey of the buried paleosol under the Lyukas mound in Hungary. Eurasian Soil Sci. 39 (SI), S133-S140. https://doi.org/10.1134/ S1064229306130217. Barczi, A., Golyeva, A.A., Petó, A., 2009. Palaeoenvironmental reconstruction of Hungarian kurgans on the basis of the examination of palaeosoils and phytolith analysis. Quat. Int. 193 (1-2), 49-60. https://doi.Org/10.1016/j. quaint.2007.10.025. Bentley, R.A. et al. (2012) 'Community differentiation and kinship among Europe's first farmers', Proceedings of the National Academy of Sciences of the United States of America, 109(24), pp. 9326-9330. doi: 10.1073/pnas.lll3710109. Bezdán, M. (2010) 'Characteristics of the Flow Regime Along the Regulated Tisza River Reach Downstream of Tiszafiired', Journal of Env. Geogr, 3, pp. 25-30 (Accessed: 29 August 2019). Bickle, P. et al. (eds.) (2017) The Neolithic of Europe: Papers in honour of Alasdair Whittle. Oxford: Oxbow books. 23 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Bille, M. and S0rensen, T.F. (2019) 'Material Culture', in Löpez-Varela, S. (ed.) The encyclopedia of archaeological sciences. Chichester, West Sussex: Wiley Blackwell, pp. 1-5. Birkäs, M., et al, 2004. Tillage effects on compaction, earthworms and other soil quality indicators in Hungary. Soil Tillage Res. 78 (2), 185-196. https://doi.Org/10.1016/j. still.2004.02.006. Bogaard, A. et al. (2013) 'Crop manuring and intensive land management by Europe's first farmers', Proceedings of the National Academy of Sciences of the United States of America, 110(31), pp. 12589-12594. doi: 10.1073/pnas.l305918110. Bogaard, A., Bending, J. and Jones, G. (2007) 'Archaeobotantical evidence for plant husbandry and use', in Whittle, A.W.R. (ed.) The Early Neolithic on the Great Hungarian Plain: Investigations of the Koros culture site of Ecsegfalva 23, County Bekes. (Varia archaeologica Hungarica, 21). Budapest: Archaeological Inst, of the HAS, pp. 421-445. Bonsall, C, et al., 2002. Climate, floods and river gods. Before Farming 3-4, 1-15. https://doi.Org/10.3828/bfarm.2002.3-4.2. Bozän, C, et al., 2018. Integrated spatial assessment of inland excess water hazard on the Great Hungarian Plain. Land Degrad. Dev. 29 (12), 4373-4386. https://doi.org/ 10.1002/ldr.3187. Brounen, F.T.S., Drenth, E. and Schut, P.A.C. (2010) 'La Hoguette north of the Rhine', in Vanmontfort, B. et al. (eds.) Pots, farmers and foragers: Pottery traditions and social interaction in the earliest Neolithic of the Lower Rhine Area. (Archaeological studies Leiden University, 20). Leiden: Leiden University Press, 95-104. Catt, J.A., 2001. The agricultural importance of loess. Earth Sei. Rev. 54, 213-229. Csengeri, P., 2010. 'Settlements of the Biikk Culture from Hernäd Valley. North-Eastern Hungary', Archeometriai Muhely 4, 227-237. Demeny, A. et al. (2013) 'Mid-Holocene climate conditions and moisture source variations based on stable H, C and O isotope compositions of speleothems in Hungary', Quaternary International (293), pp. 150-156 (Accessed: 26 August 2019). Depaer men tier, M.L.C., et al., 2020a. Neolithic land-use, subsistence, and mobility patterns in Transdanubia: a multiproxy isotope and environmental analysis from Alsonyek - Bätaszek and Morägy - Tüzködomb. J. Archaeolog. Sei.: Rep. 33 (17, pp). https://doi.Org/10.1016/j.jasrep.2020.102529. Depaer men tier, M.L.C. et al. (2020) 'Tracing mobility patterns throughout the 6-5th millennia BC in Transdanubia and the Great Hungarian Plain with strontium and oxygen stable isotope analyses (in revision)'. Dobos, E. et al. (2000) 'Use of combined digital elevation model and satellite radiometric data for regional soil mapping', Geoderma, 97, pp. 367-391 (Accessed: 28 August 2019). Domboroczki, L. and Raczky, P. (2010) 'Excavations at Ibräny-Nagyerdö and the northernmost distribution of the Koros culture in Hungary', in Koztowski, J.K. and Raczky, P. (eds.) Neolithization of the Carpathian Basin: Northernmost distribution of the Stareevo/Körös culture : papers presented on the symposium organized by the EU project FEPRE (the formation or Europe: prehistoric population dynamics and the roots of socio-cultural diversity). Krakow, pp. 191-218. Doppler, T., et al., 2017. Landscape opening and herding strategies: Carbon isotope analyses of herbivore bone collagen from the Neolithic and Bronze Age lakeshore site of Zurich-Mozartstrasse, Switzerland. Quaternary Int. 436, 18-28. https://doi.org/ 10.1016/j.quaint.2015.09.007. Eckmeier, E., et al, 2007. Pedogenesis of Chernozems in Central Europe — a review. Geoderma 139 (3-4), 288-299. https://doi.Org/10.1016/j.geoderma.2007.01.009. Eichmann, W.R., Kertesz, R. and Marton, T. (2010) 'Mesolithic in the LBK heartland of Transdaubia, Hungary', in Gronenborn, D. and Petrasch, J. (eds.) Die Neolithisierung Mitteleuropas: Internationale Tagung, Mainz 24. bis 26. Juni 2005 = The Spread of the Neolithic to Central Europe ; International Symposium, Mainz 24 June-26 June 2005. (RGZM-Tagungen, 4). Mainz: Verlag des Römisch-Germanischen Zentralmuseums, pp. 211-233. Feurdean, A., 2005. Holocene forest dynamics in northwestern Romnania. Holocene 15 (3), 435-446. https://doi.org/10.1191/0959683605hl803rp. Feurdean, A., et al., 2020. The transformation of the forest steppe in the lower Danube Plain of south-eastern Europe: 6000 years of vegetation and land use dynamic. Biogeosciences 32. https://doi.org/10.5194/bg-2020-239. Gardner, A.R., 2002. Neolithic to Copper Age woodland impacts in northeast Hungary? Evidence from the pollen and sediment chemistry records. Holocene 12 (5), 541-553. https://doi.org/10.1191/0959683602M561rp. Garnett, A., 1945. The Loess Regions of Central Europe in Prehistoric Times. Geogr. J. 106 (3/4), 132-143. Gerlach, R. and Eckmeier, E. (2012) 'Prehistoric Land Use and Its Impact on Soil Formation since Early Neolithic. Examples from the Lower Rhine Area', in Bebermeier, W. et al. (eds.) Landscape Archaeology Conference (LAC 2012). eTopoi. (Journal for Ancient Studies, Special Volume 3). Berlin, pp. 11-16 (Accessed: 4 December 2019). Gerling, C, et al., 2017. High-resolution isotopic evidence of specialised cattle herding in the European Neolithic. PLoS ONE 12 (7), e0180164. https://doi.org/10.1371/ journal, pone. 0180164. Giblin, J.I., et al., 2013. Strontium isotope analysis and human mobility during the Neolithic and Copper Age: a case study from the Great Hungarian Plain. J. Archaeol. Sei. 40 (1), 227-239. https://doi.Org/10.1016/j.jas.2012.08.024. Gillings, M., 1995. Flood dynamics and settlement in the Tisza valley of north-east Hungary: GIS and the Upper Tisza project. In: Lock, G., Stancic, Z. (Eds.), Archaeology and Geographical Information Systems: A European Perspective. Taylor & Francis, London, pp. 67-84. Gillings, M. (2007) 'The Ecsegfalva Landscape: Affordance and Inhabitation', in Whittle, A.W.R. (ed.) The Early Neolithic on the Great Hungarian Plain: Investigations of the Koros culture site of Ecsegfalva 23, County Bekes. (Varia archaeologica Hungarica, 21). Budapest: Archaeological Inst, of the HAS, pp. 31-46. Groenhuijzen, M.R. (2019) 'Palaeogeographie-Analysis Approaches to Transport and Settlement in the Dutch Part of the Roman Limes', in Verhagen, P., Joyce, J. and Groenhuijzen, M.R. (eds.) Finding the limit of the Limes. (Computational Social Sciences) [Place of publication not identified]: SPRINGER NATURE, pp. 251-269 (Accessed: 29 November 2019). Gronenborn, D. (1999) 'A Variation on a Basic Theme: The Transition to Farming in Southern Central Europe', Journal of world prehistory, 13(2) (Accessed: 7 December 2019). Guida, R.J., et al., 2015. Strategic floodplain reconnection for the Lower Tisza River, Hungary: opportunities for flood-height reduction and floodplain-wetland reconnection. J. Hydrol. 521, 274-285. https://doi.0rg/lO.lOl6/j. jhydrol.2014.11.080. Güimil-Farina, A., Parcero-Oubina, C, 2015. "Dotting the joins": a non-reconstructive use of Least Cost Paths to approach ancient roads. The case of the Roman roads in the NW Iberian Peninsula. J. Archaeol. Sei. 54, 31-44. https://doi.0rg/lO.lOl6/j. jas.2014.11.030. Haak, W., et al., 2005. Ancient DNA from the first European farmers in 7500-year-old Neolithic sites. Science (New York, N.Y.) 310 (5750), 1016-1018. https://doi.org/ 10.1126/science. 111872 5. Haak, W., et al., 2015. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 522 (7555), 207-211. https://doi.org/ 10.1038/naturel4317. Hammes, K. et al. (2007) 'Comparison of quantification methods to measure fire-derived (black/elemental) carbon in soils and sediments using reference materials from soil, water, sediment and the atmosphere', Global Biogeochemical Cycles, 21(3), n/a-n/a. doi: 10.1029/2006GB002914. Harding, A., 2015. Salt exploitation in the later prehistory of the Carpathian Basin. Documenta Praehistorica 42, 211-217. Harris, O.J.T., 2017. Assemblages and Scale in Archaeology. Cambridge Archaeol. J. 27 (1), 127-139. https://doi.org/10.1017/S0959774316000597. Hedges, R., et al., 2013. The supra-regional perspective. In: Bickle, P., Whittle, A. (Eds.), The first farmers of central Europe: Diversity in LBK lifeways. Oxbow Books and the David Brown Book Company, Oxford. Hertelendi, E., et al., 1995. Re-evaluation of the neolithic in eastern hungary based on calibrated radiocarbon dates. Radiocarbon 37 (2), 239-244. https://doi.org/ 10.1017/S0033822200030691. Hertelendi, E., Horväth, F., 1992. Radiocarbon chronology of late neolithic settlements in the Tisza-Maros Region, Hungary. Radiocarbon 34 (3), 859-866. https://doi.org/ 10.1017/S0033822200064171. Hertelendi, E., Siimegi, P., Szöör, G., 1992. Geochronologic and paleoclimatic characterization of quaternary sediments in the Great Hungarian Plain. Radiocarbon 34 (3), 833-839. https://doi.org/10.1017/S0033822200064146. Heyvaert, V.M., et al, 2012. The role of human interference on the channel shifting of the Karkheh River in the Lower Khuzestan plain (Mesopotamia, SW Iran). Quat. Int. 251, 52-63. https://doi.Org/10.1016/j.quaint.2011.07.018. Heyvaert, V.M., Walstra, J., 2016. The role of long-term human impact on avulsion and fan development. Earth Surf. Proc. Land. 41 (14), 2137-2152. https://doi.org/ 10.1002/esp.4011. Hijmans, R.J., et al., 2005. Very high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25 (15), 1965-1978. https://doi.org/10.1002/joc.1276. Howey, M.C.L., 2011. Multiple pathways across past landscapes: circuit theory as a complementary geospatial method to least cost path for modeling past movement. J. Archaeol. Sei. 38 (10), 2523-2535. https://doi.Org/10.1016/j.jas.2011.03.024. Howey, M.C.L., Brouwer Burg, M., 2017. Assessing the state of archaeological GIS research: unbinding analyses of past landscapes. J. Archaeol. Sei. 84, 1-9. https:// doi.org/10.1016/j.jas.2017.05.002. Jakab, G. et al. (2009) 'Palaeoc lima tic signals and anthropogenic disturbances from the peatbog at Nagybärkäny (North Hungary)', in Buczkö, K. et al. (eds.) Palaeolimnological Proxies as Tools of Environmental Reconstruction in Fresh Water. (Developments in Hydrobiology, 208). Dordrecht: Springer Science+Business Media B.V, pp. 87-106. Jakucs, J., et al., 2016. Between the Vinca and Linearbandkeramik Worlds: the Diversity of Practices and Identities in the 54th-53rd Centuries cal BC in Southwest Hungary and Beyond. J. World Prehistory 29 (3), 267-336. https://doi.org/10.1007/sl0963- 016-9096-x. Kabala, C, et al, 2019. Origin, age and transformation of Chernozems in northern Central Europe - New data from Neolithic earthen barrows in SW Poland. CATENA 180, 83-102. https://doi.Org/10.1016/j.catena.2019.04.014. Kalicz, N. (1994) 'A dunäntüli (közep-euröpai) vonaldiszes kerämia legidösebb leletei es a korai Vinca kultura - Die ältesten Funde der transdanubischen (mitteleuropäischen) Linienbandkeramik und die frühe Vinca-Kultur', in Lörinczy, G. (ed.) A kökortöl a közepkorig. Tanulmänyok Trogmayer Otto 60. születesnapjära. - Von der Steinzeit bis zum Mittelalter. Studien zu 60. Geburtstag von Otto Trogmayer. Szeged, pp. 67-85. Kalicz, N. and Makkay, J. (1977) Die Linienbandkeramik in der Großen Ungarischen Tiefebene. (Studia archaeologica, 7). Budapest: Akad. Kiadö. Kalicz, N. and Raczky, P. (1987) 'The Late Neolithic of the Tisza Region. A survey of recent archaeological research.', in Tälas, L. (ed.) The late Neolithic of the Tisza region: A survey of recent excavations and their findings. Budapest: Szolnok, pp. 11-19. Kasse, C, et al., 2010. Fluvial style changes during the last glacial-interglacial transition in the middle Tisza valley (Hungary). Proc. Geol. Assoc. 121 (2), 180-194. https:// doi.org/10.1016/j.pgeola.2010.02.005. Kempf, M., 2019a. Paradigm and pragmatism: GIS-based spatial analyses of Roman infrastructure networks and land-use concepts in the Upper Rhine Valley. Geoarchaeology 74 (285), 1-12. https://doi.org/10.1002/gea.21752. 24 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Kempf, M., 2019b. The application of GIS and satellite imagery in archaeological land-use reconstruction: A predictive model? J. Archaeolog. Sci.: Rep. 25, 116-128. https://doi.Org/10.1016/j.jasrep.2019.03.035. Kempf, M., 2020. Modeling multivariate landscape affordances and functional ecosystem connectivity in landscape archeology. Archaeol. Anthropol. Sci. 12 (8), 1-21. https://doi.org/10.1007/sl2520-020-01127-w. Kempf, M. et al. (2020) 'Modelling the first strontium isotope baseline map for the Carpathian Basin (in preparation)'. Kercsmár, Z. et al. (eds.) (2015) Surface geology of Hungary: Explanatory notes to the geological map of Hungary (1:500 000). Budapest: Geological and Geophysical Institute of Hungary (Accessed: 4 August 2019). Kerényi, A., 1994. Loess Erosion on the Tokaj Big-Hill. Quat. Int. 24, 47-52. Kirby, K.J. (2004) 'A model of a natural wooded landscape in Britain as influenced by large herbivore activity', Forestry, 77(5) (Accessed: 21 August 2020). Kiss, T., et al., 2015. The evolution of the Great Hungarian Plain fluvial system - Fluvial processes in a subsiding area from the beginning of the Weichselian. Quat. Int. 388, 142-155. https://doi.Org/10.1016/j.quaint.2014.05.050. Kiss, T., Sumeghy, B., Sipos, G., 2014. Late Quaternary paleodrainage reconstruction of the Maros River alluvial fan. Geomorphology 204, 49-60. https://doi.org/10.1016/ j.geomorph.2013.07.028. Kleber, M., et al., 2003. Prehistoric Alteration of soil properties in a Central German chernozemic soil: in search of pedologie indicators for prehistoric activity. Soil Sci. 168 (4), 292. https://doi.org/10.1097/01.ss.0000064892.94869.3a. Knipl, I., Siimegi, P., 2012. Life at the interface of two distinct landscapes-relationship of humans and environment in the periphery of the Danube-Tisza Interfluve between Hajós and Császártóltés. Open Geosci. 4 (3), 1. https://doi.org/10.2478/sl3533- 011-0071-x. Knitter, D., et al., 2018. Critical physical geography in practice: landscape archaeology. In: Lave, R., Biermann, C, Lane, S.N. (Eds.), The Palgrave Handbook of Critical Physical Geography. (Palgrave handbooks). Palgrave Macmillan, Cham, pp. 179-200. Kovács, J., et al., 2013. Clay mineralogy of red clay deposits from the central Carpathian Basin (Hungary): implications for Plio-Pleistocene chemical weathering and palaeoclimate. Turk. J. Earth Sci. 22, 414-426. https://doi.org/10.3906/yer-1201-4. Krasilnikov, P.V., Michélí, E., 2009. Soil Classification of Hungary. In: Krasilnikov, P.V. (Ed.), A Handbook of Soil Terminology, Correlation and Classification. Sterling (USA), Earthscan, pp. 170-175. Kreuz, A., 2007. Closed forest or open woodland as natural vegetation in the surroundings of Linearbandkeramik settlements? Vegetation History Archaeobotany 17 (1), 51-64. https://doi.org/10.1007/s00334-007-0110-l. Kvamme, K., 2020. Analysing regional environmental relationships. In: Gillings, M., Hacigiizeller, P., Lock, G. (Eds.), Archaeological Spatial Analysis: A Methodological Guide. Routledge, New York, pp. 212-230. Laborczi, A., et al., 2016. Mapping of topsoil texture in Hungary using classification trees. J. Maps 12 (5), 999-1009. https://doi.org/10.1080/17445647.2015.1113896. Laborczi, A., et al., 2019. Comparison of soil texture maps synthetized from standard depth layers with directly compiled products. Geoderma 352, 360-372. https://doi. org/10.1016/j.geoderma.2018.01.020. Ladányi, Z., et al., 2015. Multi-indicator sensitivity analysis of climate change effects on landscapes in the Kiskunság National Park, Hungary. Ecol. Indicators 58, 8-20. https://doi.Org/10.1016/j.ecolind.2015.05.024. Lasaponara, R., Masini, N., 2006. Identification of archaeological buried remains based on the normalized difference vegetation index (NDVI) from Quickbird satellite data. IEEE Geosci. Remote Sens. Lett. 3 (3), 325-328. https://doi.org/10.1109/ LGRS.2006.871747. Lipson, M., et al., 2017. Parallel palaeogenomic transects reveal complex genetic history of early European farmers. Nature 551 (7680), 368-372. https://doi.org/10.1038/ nature24476. Lorz, C, Saile, T., 2011. Anthropogenic pedogenesis of Chernozems in Germany? - A critical review. Quat. Int. 243 (2), 273-279. https://doi.Org/10.1016/j. quaint.2010.11.022. Mádl-Szónyi, J., Tóth, J., Pogácsás, G., 2008. Soil and wetland salinization in the framework of the Danube-Tisza Interfluve hydrogeologie type section. Central Eur. Geol. 51 (3), 203-217. https://doi.Org/10.1556/CEuGeol.51.2008.3.3. Magyari, E., et al., 2001. Retarded wetland succession: anthropogenic and climatic signals in a Holocene peat bog profile from north-east Hungary. J. Ecol. 89 (6), 1019-1032. https://doi.Org/10.llll/j.1365-2745.2001.00624.x. Magyari, E. (2011) 'Late Quaternary Vegetation History in the Hortobágy Steppe and Middle Tisza Floodplain, NE Hungary', Studia bot. hung., 42, pp. 185-203 (Accessed: 31 August 2019). Magyari, E.K., et al, 2009. Radiocarbon chronology of glacial lake sediments in the Retezat Mts (South Carpathians, Romania): a window to Late Glacial and Holocene climatic and paleoenvironmental changes. Central Eur. Geol. 52 (3-4), 225-248. https://doi.Org/10.1556/CEuGeol.52.2009.3-4.2. Magyari, E.K., et al, 2010. Holocene persistence of wooded steppe in the Great Hungarian Plain. J. Biogeogr. 37 (5), 915-935. https://d0i.0rg/l O.llll/j.1365- 2699.2009.02261.x. Maher, B.A., et al., 2003. Holocene loess accumulation and soil development at the western edge of the Chinese Loess Plateau: implications for magnetic proxies of palaeorainfall. Quat. Sci. Rev. 22 (5-7), 445-451. https://doi.org/10.1016/S0277-3791(02)00188-9. Mathieson, I., et al., 2015. Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528 (7583), 499-503. https://doi.org/10.1038/naturel6152. Michéli, E., et al., 2006. Classification of the Major Soils of Hungary and their Correlation with the World Reference Base for Soil Resources (WRB). Agrokémia és Talajtan 55 (1), 19-28. https://doi.Org/10.1556/Agrokem.55.2006.l.3. Molnár, S. and Siimegi, P. (2007) 'A long history of the Kiri-Tó meander', in Whittle, A. W.R. (ed.) The Early Neolithic on the Great Hungarian Plain: Investigations of the Koros culture site of Ecsegfalva 23, County Békés. (Varia archaeologica Hungarica, 21). Budapest: Archaeological Inst, of the HAS, pp. 47-65. Molnár, Z. et al. (2012) 'Past Trends, Present State and Future Prospects of Hungarian Forest-Steppes', in Werger, M.J.A. and van Staalduinen, M.A. (eds.) Eurasian Steppes. Ecological Problems and Livelihoods in a Changing World. (Plant and Vegetation, 6). Dordrecht: Springer Netherlands, pp. 209-252. Montandon, L., Small, E., 2008. The impact of soil reflectance on the quantification of the green vegetation fraction from NDVI. Remote Sens. Environ. 112 (4), 1835-1845. https://doi.Org/10.1016/j.rse.2007.09.007. Moskal-del Hoyo, M., et al., 2018. The character of the Atlantic oak woods of the Great Hungarian Plain. Quat. Int. 463, 337-351. https://doi.0rg/lO.lOl6/j. quaint.2017.02.029. Müller, J., Diachenko, A., 2019. Tracing long-term demographic changes: the issue of spatial scales. PLoS ONE 14 (1), e0208739. https://doi.org/10.1371/journal. pone.0208739. Nádor, A., et al., 2011. Late Quaternary (Weichselian) alluvial history and neotectonic control on fluvial landscape development in the southern Koros plain, Hungary. Palaeogeogr. Palaeoclimatol. Palaeoecol. 299 (1-2), 1-14. https://doi.org/10.1016/ j.palaeo.2010.09.018. Nagy, L, Ligetvári, F. and Schweitzer, F. (2010) 'Tisza River Valley: future prospects', Hungarian geographical bulletin (59), pp. 361-370 (Accessed: 5 December 2019). Nakoinz, O. and Knitter, D. (2016) Modelling Human Behaviour in Landscapes: Basic Concepts and Modelling Elements. (Quantitative Archaeology and Archaeological Modelling). Cham: Springer International Publishing. Novothny, A., Frechen, M., Horváth, E., 2010. Luminescence dating of periods of sand movement from the Gödöllö Hills, Hungary. Geomorphology 122 (3-4), 254-263. https://doi.Org/10.1016/j.geomorph.2010.04.013. Nyári, D., et al., 2014. Environmental Changes in Historical Times Near Apostag on the Danube-Tisza Interfluve, Hungary (A Complex Research Based on Archaeological Excavation and Geo morphological Investigations). J. Environ. Geogr. 7 (1-2), 39-47. https://doi.org/10.2478/jengeo-2014-0005. Ochs, K., et al., 2020. The potential natural vegetation of large river floodplains - from dynamic to static equilibrium. J. Hydro-environ. Res. 30, 71-81. https://doi.org/ 10.1016/j.jher.2020.01.005. Oross, K., Bánffy, E., 2009. Three successive waves of Neolithisation: LBK development in Transdanubia. Documenta Praehistorica 36, 175. https://doi.org/10.4312/ dp.36.11. Pásztor, L., et al., 2012. Compilation of 1:50,000 scale digital soil maps for Hungary based on the digital Kreybig soil information system. J. Maps 8 (3), 215-219. https://doi.org/10.1080/17445647.2012.705517. Pásztor, L., et al., 2014a. Application of legacy soil data in digital soil mapping for the elaboration of novel, countrywide maps of soil conditions. Agrokémia és Talajtan 63 (1), 79-88. https://doi.Org/10.1556/Agrokem.63.2014.l.9. Pásztor, L. et al. (2014) 'Elaboration of novel, countrywide maps for the satisfaction of recent demands on spatial, soil related information in Hungary', in Arrouays, D. et al. (eds.) GlobalSoilMap: Basis of the global spatial soil information system : proceedings of the 1st GlobalSoilMap conference, Orleans, France, 7-9 October 2013. Boca Raton: CRC Presss, pp. 207-212. Pásztor, L., et al., 2015a. Compilation of novel and renewed, goal oriented digital soil maps using geostatistical and data mining tools. Hungarian Geogr. Bull. 64 (1), 49-64. https://doi.Org/10.15201/hungeobull.64.l.5. Pásztor, L., et al., 2015b. Spatial risk assessment of hydrological extremities: Inland excess water hazard, Szabolcs-Szatmár-Bereg County, Hungary. Journal of Maps 11 (4), 636-644. https://doi.org/10.1080/17445647.2014.954647. Pásztor, L., et al., 2016. Integrated spatial assessment of wind erosion risk in Hungary. Nat. Hazards Earth Syst. Sci. 16 (11), 2421-2432. https://doi.org/10.5194/nhess- 16-2421-2016. Pásztor, L., et al., 2018. Compilation of a national soil-type map for Hungary by sequential classification methods. Geoderma 311, 93-108. https://doi.org/10.1016/ j.geoderma.2017.04.018. Pásztor, L., Szabó, J. and Bakacsi, Z. (2002) 'GIS Processing of Large-scale Soil Maps in Hungary', Agrokémia és Talajtan, 51, pp. 273-282 (Accessed: 28 August 2019). Petrovszki, J., Timár, G., 2010. Channel sinuosity of the Koros River system, Hungary/ Romania, as possible indicator of the neotectonic activity. Geomorphology 122 (3-4), 223-230. https://doi.Org/10.1016/j.geomorph.2009.ll.009. Pinke, Z., Lövei, G.L., 2017. Increasing temperature cuts back crop yields in Hungary over the last 90 years. Glob. Change Biol. 23 (12), 5426-5435. https://doi.org/ 10.1111/gcb.l3808. Preston, CM. and Schmidt, M.W.I. (2006) 'Black (pyrogenic) carbon: a synthesis of current knowledge and uncertainties with special consideration of boreal regions', Biogeosciences, 3, pp. 397-420 (Accessed: 21 August 2020). Raczky, P. et al. (2005) 'Zwischen Himmel und Erde - Polgar-Csöszhalom, eine Siedlung in Ostungarn', in Daim, F. and Neubauer, W. (eds.) "Zeitreise Heldenberg" -Geheimisvolle Kreisgräben: Katalog zur Niederösterreichischen Landesausstellung 2005. (Katalog des NÖ Landesmuseums, N.F., 459). Horn: Berner, pp. 203-209. Raczky, P. and Anders, A. (2009) 'Settlement history of the Middle Neolithic in the Polgár micro-region (The development of the Alföld Linearband Pottery in the Upper Tisza Region, Hungary)', in Koztowski, J.K. (ed.) Interactions between different models of neolithization North of the Central European agro-ecological barrier: Papers presented on the symposium organized by the EU Project FEPRE (The Formation of Europe : Prehistoric Population Dynamics and the Roots of Socio-Cultural Diversity) = Interakcje pomi^dzy rôznymi modelami neolityzacji na polnoc od srodkowoeuropejskiej granicy agro-ekologicznej. (Prace Komisji Prehistorii Karpat, 5). Krakow: Polska Akadémia Umiejetnosci, pp. 31-50. 25 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 Rivollat, M., et al., 2016. Ancient mitochondrial DNA from the middle neolithic necropolis of Obernai extends the genetic influence of the LBK to west of the Rhine. Am. J. Phys. Anthropol. 161 (3), 522-529. https://doi.org/10.1002/ajpa.23055. Roberts, N., Rosen, A., 2009. Diversity and complexity in early farming communities of Southwest Asia: new insights into the economic and environmental basis of Neolithic Catalhöyük. Curr. Anthropol. 50 (3), 393-402. https://doi.org/10.1086/598606. Roper, D.C. (1979) 'The Method and Theory of Site Catchment Analysis: A Review', Advances in Archaeological Method and Theory, 2, pp. 119-140 (Accessed: 6 November 2019). Salisbury, R.B., 2013. Interpolating geochemical patterning of activity zones at Late Neolithic and Early Copper Age settlements in eastern Hungary. J. Archaeol. Sei. 40 (2), 926-934. https://doi.Org/10.1016/j.jas.2012.10.009. Saqalli, M., et al, 2014. Revisiting and modelling the woodland farming system of the early Neolithic Linear Pottery Culture (LBK), 5600-4900 b.c. Vegetation History Archaeobotany 23 (SI), 37-50. https://doi.org/10.1007/s00334-014-0436-4. Schmidt, M.W.I, et al. (1999) 'Charred organic carbon in German chernozemic soils', European Journal of Soil Science (50), pp. 351-365 (Accessed: 21 August 2020). Schmidt, M.W.I., Noack, A.G., 2000. Black Carbon in soils and sediments: analysis, distribution, implications, and current challenges. Global Biogeochem. Cycles 14 (3), 777-793. Schmidt, M.W.I., Skjemstad, J.O. and Jäger, C. (2002) 'Carbon isotope geochemistry and nanomorphology of soil black carbon: Black chernozemic soils in central Europe originate from ancient biomass burning', Global Biogeochemical Cycles, 16(4), 70-1-70-8. doi: 10.1029/2002GB001939. Schofield, R., Thomas, D.S.G., Kirkby, M.J., 2001. Causal processes of soil salinization in Tunisia, Spain and Hungary. Land Degrad. Dev. 12 (2), 163-181. https://doi.org/ 10.1002/ldr.446. Schulze, M.-S., 2000. Patterns of growth and stagnation in the late nineteenth century Habsburg economy. Eur. Rev. Economic History 4 (3), 311-340. https://doi.org/ 10.1017/S1361491600000095. Schweitzer, F. (2009) 'Strategy or Disaster : Flood Prevention Related Issues and Actions in the Tisza River Basin', Hungarian geographical bulletin (58(1)), pp. 3-17 (Accessed: 5 December 2019). Sherratt, A. (1983) 'The Development of Neolithic and Copper Age Settlement in the Great Hungarian Plain: Part II: Site Survey and Settlement Dynamics', Oxford Journal of Archaeology, 2(1), pp. 13-41 (Accessed: 6 November 2019). Sielmann, B. (1972) 'Die frühneolithische Besiedlung Mitteleuropas', in Schwabedissen, H. (ed.) Die Anfänge des Neolithikums vom Orient bis Nordeuropa Westliches Mitteleuropa. Köln: Böhlau, pp. 1-65. Sipos, G., Kiss, T., Toth, O., 2016. Constraining the age of floodplain levels along the lower section of River Tisza, Hungary. J. Environ. Geogr. 9 (1-2), 39-44. https://doi. org/10.1515/jengeo-2016-0006. Skejmstad, J.O., et al., 1997. Characterization of soil organic matter by solid-state 13C NMR spectroscopy. In: Cadisch, G., Giller, K.E. (Eds.), Driven by Nature: Plant Litter Quality and Decomposition. CAB International, Wallingford, pp. 253-271. Smalley, I., Markovic, S.B., Svircev, Z., 2011. Loess is [almost totally formed by] the accumulation of dust. Quat. Int. 240 (1-2), 4-11. https://doi.Org/10.1016/j. quaint.2010.07.011. Stadler, P. and Ruttkay, E. (2007) 'Absolute Chronology of the Mora vi an-Eastern-Austrian Group (MOG) of the Painted Pottery (Lengyel-Culture) Based on New Radiocarbon Dates from Austria', in Koztowski, J.K. and Raczky, P. (eds.) The Lengyel, Polgär and related cultures in the Middle/Late Neolithic in Central Europe: Symposium ... at the Polish Academy of Arts and Sciences in Krakow, between March 7th and 12th, 2006. Krakow, pp. 117-146. Starling, N.J. (1983) 'Neolithic Settlement Patterns in Central Germany', Oxford Journal of Archaeology (2(1)), pp. 1-11 (Accessed: 4 December 2019). Stouthamer, E., Berendsen, H.J.A., 2000. Factors controlling the holocene avulsion history of the Rhine-Meuse Delta (The Netherlands). J. Sediment. Res. 70 (5), 1051-1064. https://doi.org/10.1306/033000701051. Strouhalovä, B., et al, 2019. Assessing the vegetation history of European Chernozems through qualitative near infrared spectroscopy. Quaternaire 30 (3), 227-241. Siimegi, P., et al., 2013a. Responses of terrestrial ecosystems to Dansgaard-Oeshger cycles and Heinrich-events: a 28,000-year record of environmental changes from SE Hungary. Quat. Int. 293, 34-50. https://doi.Org/10.1016/j.quaint.2012.07.032. Siimegi, P., et al., 2013b. The Late Quaternary paleoecology and environmental history of Hortobägy, a unique mosaic alkaline steppe from the heart of the Carpathian Basin. In: Morales Prieto, M.B., Traba Diaz, J. (Eds.), Steppe ecosystems: Biological diversity, management and restoration. (Environmental research advances). Nova Publishers, New York, pp. 165-193. Siimegi, P., et al., 2015. First radiocarbon dated paleoecological data from the freshwater carbonates of the Danube-Tisza Interfluve. Open Geosciences 7 (1), 239. https://doi. org/10.1515/geo-2015-0003. Siimegi, P., et al., 2019. High-resolution proxy record of the environmental response to climatic variations during transition MIS3/MIS2 and MIS2 in Central Europe: the loess-paleosol sequence of Katymär brickyard (Hungary). Quat. Int. 504, 40-55. https://doi.Org/10.1016/j.quaint.2018.03.030. Siimegi, P., Gulyäs, S. and Persaits, G. (2013) 'The Geoarchaeological Evolution of the Loess-covered Alluvial Island of Polgär and its Role in Shaping Human Settlement Strategies', in Anders, A. and Kulcsär, G. (eds.) Moments in time: Papers presented to Päl Raczky on his 60th birthday. (Prehistoric studies = Osregeszeti tanulmänyok, 1). Budapest: L'Harmattan, pp. 1-12 (Accessed: 10 December 2019). Siimegi, P., Persaits, G., Gulyäs, S., 2012. Woodland-Grassland Ecotonal Shifts in Environmental Mosaics: lessons Learnt from the Environmental History of the Carpathian Basin (Central Europe) During the Holocene and the Last Ice Age Based on Investigation of Paleobotanical and Mollusk Remains. In: Myster, R.W. (Ed.), Ecotones between forest and grassland. Springer, New York, NY, pp. 17-57. Szécsényi-Nagy, A. et al. (2014) 'Ancient DNA supports a homogeneous maternal gene pool in the 6th millennium BC Hungary and in Central European LBK', in Whittle, A. W.R. and Bickle, P. (eds.) Early farmers: The view from archaeology and science. (Proceedings of the British Academy, 198). Oxford: Oxford Univ. Press, pp. 71-93. Szécsényi-Nagy, A. (2015) Molecular genetic investigation of the Neolithic population history in the western Carpathian Basin. PhD thesis. Szécsényi-Nagy, A., et al., 2015. Tracing the genetic origin of Europe's first farmers reveals insights into their social organization. Proc. Biol. Sci. 282 https://doi.org/ 10.1098/rspb.2015.0339. Szilassi, P., et al., 2006. Impacts of historical land use changes on erosion and agricultural soil properties in the Kali Basin at Lake Balaton, Hungary. CATENA 68 (2-3), 96-108. https://doi.Org/10.1016/j.catena.2006.03.010. Thiele-Bruhn, S., et al., 2014. Chernozem properties of Black Soils in the Baltic region of Germany as revealed by mass-spectrometric fingerprinting of organic matter. Geoderma 213, 144-154. https://doi.Org/10.1016/j.geoderma.2013.07.036. Timár, G., Siimegi, P., Horváth, F., 2005. Late Quaternary dynamics of the Tisza River: evidence of climatic and tectonic controls. Tectonophysics 410 (1-4), 97-110. https://doi.Org/10.1016/j.tecto.2005.06.010. Tóth, T., et al., 2001. Use of digitalized hydrogeologie al maps for evaluation of salt-affected soils of large areas. Arid Land Res. Manage. 15 (4), 329-346. https://doi. org/10.1080/153249801753127624. Uj, B., et al., 2016. Wetland habitats of the Kis-Sárrét 1860-2008 (Kórós-Maros National Park, Hungary). J. Maps 12 (2), 211-221. https://doi.org/10.1080/ 17445647.2014.1001799. Ujházy, K., Gábris, G., Frechen, M., 2003. Ages of periods of sand movement in Hungary determined through luminescence measurements. Quat. Int. Ill (1), 91-100. https://doi.org/10.1016/Sl 040-6182(03)00017-X. Ujházy, N., Biró, M., 2018. The 'Cursed Channel': Utopian and dystopian imaginations of landscape transformation in twentieth-century Hungary. J. Historical Geogr. 61, 1-13. https://doi.Org/10.1016/j.jhg.2018.01.001. Ullah, I.I.T., 2011. A GIS method for assessing the zone of human-environmental impact around archaeological sites: a test case from the Late Neolithic of Wadi Ziqláb, Jordan. J. Archaeol. Sci. 38 (3), 623-632. https://doi.0rg/lO.lOl6/j. jas.2010.10.015. van der Zee, S.E.A.T.M. et al. (2017) 'Irrigation and Drainage in Agriculture: A Salinity and Environmental Perspective', in Kulshreshtha, S. and Elshorbagy, A. (eds.) Current Perspective on Irrigation and Drainage: InTech, pp. 1-21 (Accessed: 24 August 2019). van Dinter, M., 2013. The Roman Limes in the Netherlands: how a delta landscape determined the location of the military structures. Netherlands J. Geosci. - Geologie en Mijnbouw 92 (1), 11-32. https://doi.org/10.1017/S0016774600000251. van Lanen, R.J., et al., 2015a. Best travel options: modelling Roman and early-medieval routes in the Netherlands using a multi-proxy approach. J. Archaeolog. Sci.: Rep. 3, 144-159. https://doi.Org/10.1016/j.jasrep.2015.05.024. van Lanen, R.J., et al., 2015b. Finding a way: modeling landscape prerequisites for roman and early-medieval routes in the Netherlands. Geoarchaeology 30 (3), 200-222. https://doi.org/10.1002/gea.21510. Vanmontfort, B. et al. (eds.) (2010) Pots, farmers and foragers: Pottery traditions and social interaction in the earliest Neolithic of the Lower Rhine Area. Leiden: Leiden University Press (Archaeological studies Leiden University, 20). Várallay, G. (1989) 'Soil degradation processes and their control in Hungary', Land Degradation and Rehabilitation, 1, pp. 171-188 (Accessed: 28 August 2019). Varga, G. (2011) 'Similarities among the Plio-Pleistocene terrestrial aeolian dust deposits in the World and in Hungary', Quaternary International (234), pp. 98-108 (Accessed: 26 August 2019). Vargha, A., Delaney, H.D., 2000. A critique and improvement of the CL common language effect size statistics of McGraw and Wong. J. Educ. Behav. Statistics 25 (2), 101-132. Verhagen, P. (2019) 'Predictive Modeling', in López-Varela, S. (ed.) The encyclopedia of archaeological sciences. Chichester, West Sussex: Wiley Blackwell, pp. 1-3 (Accessed: 29 November 2019). Verhulst, N., et al., 2009. Using NDVI and soil quality analysis to assess influence of agronomic management on within-plot spatial variability and factors limiting production. Plant Soil 317 (1-2), 41-59. https://doi.org/10.1007/slll04-008-9787-x. Vita-Finzi, C, et al, 1970. Prehistoric economy in the mount carmel area of palestine: site catchment analysis. Proc. Prehist. Soc 36, 1-37. https://doi.org/10.1017/ S0079497X00013074. Volkmann, A. (2018) 'Methods and Perspectives of Geoarchaelogical Site Catchment Analysis: Identification of Palaeoclimate Indicators in the Oder Region from the Iron to Middle Ages', in Siart, C, Forbriger, M. and Bubenzer, O. (eds.) Digital geoarchaeology: New techniques for interdisciplinary human-environmental research. (Natural Science in Archaeology). Cham, Switzerland: Springer, pp. 27-44 (Accessed: 6 November 2019). Vysloužilová, B., et al., 2015. Investigation of vegetation history of buried chernozem soils using near-infrared spectroscopy (NIRS). Quat. Int. 365, 203-211. https://doi. org/10.1016/j.quaint.2014.07.035. Wallace, M.P., et al., 2015. Stable carbon isotope evidence for neolithic and bronze age crop water management in the Eastern Mediterranean and Southwest Asia. PLoS ONE 10 (6), e0127085. https://doi.org/10.1371/journal.pone.0127085. Waltner, I., et al., 2018. Evaluating the new soil erosion map of Hungary-a semiquantitative approach. Land Degrad. Dev. 29 (4), 1295-1302. https://doi.org/ 10.1002/ldr.2916. Weaverdyck, E.J.S. (2019) 'The Role of Forts in the Local Market System in the Lower Rhine: Towards a Method of Multiple Hypothesis Testing Through Comparative Modelling', in Verhagen, P., Joyce, J. and Groenhuijzen, M.R. (eds.) Finding the 26 M. Kempf Journal of Archaeological Science: Reports 34 (2020) 102637 limit of the Limes. (Computational Social Sciences) [Place of publication not identified]: SPRINGER NATURE, pp. 165-190 (Accessed: 29 November 2019). Whittle, A., et al., 2002. In the beginning : new radiocarbon dates for the Early Neolithic in northern Serbia and south-east Hungary. Antaeus 25, 63-118. Whittle, A.W.R. (ed.) (2007) The Early Neolithic on the Great Hungarian Plain: Investigations of the Koros culture site of Ecsegfalva 23, County Bekes. Budapest: Archaeological Inst, of the HAS (Varia archaeologica Hungarica, 21). Willis, K.J. et al. (1997) 'Does soil change cause vegetation change or vice versa? A temporal perspective from Hungary', Ecology, 78(3), pp. 740-750 (Accessed: 27 August 2019). Willis, K.J., Rudner, E., Siimegi, P., 2000. The full-glacial forests of Central and Southeastern Europe. Quat. Res. 53 (2), 203-213. https://doi.org/10.1006/ qres. 1999.2119. Zalai-Gaäl, I. (1990) 'Neue Daten zur Erforschung der spätneolithisehen Schanzwerke im südlichen Transdanubien', in Zalai-Müzeum 2 (ed.) Die Fragen des Neolithikums und der Kupferzeit: Archäologische Konferenz des Komitates Zala und Niederösterreichs. Keszthely, pp. 31-48. 27