Digital Mapping of Habitat for Plant Communities Based on Soil Functions: a Case Study in the Virgin Forest-Steppe of Russia

Home , Chernozem, Digital mapping, Pochvovedenie, Soil functions

Article Digital Mapping of Habitat for Plant Communities Based on Soil Functions: A Case Study in the Virgin Forest-Steppe of Russia

Nikolai Lozbenev 1,* , Maria Smirnova 1,2 , Maxim Bocharnikov 2 and Daniil Kozlov 1

1 V.V. Dokuchaev Soil Science Institute, Pyzhyovskiy lane 7 building 2, Moscow 109017, Russia; [email protected] (M.S.); [email protected] (D.K.) 2 Lomonosov Moscow State University, Leninskiye Gory, 1, Moscow 119234, Russia; [email protected] * Correspondence: [email protected]; Tel.: +7-909-966-96-24

Received: 14 December 2018; Accepted: 6 March 2019; Published: 9 March 2019

Abstract: The spatial structure of the habitat for plant communities based on soil functions in virgin forest-steppe of the Central Russian Upland is the focus of this study. The objectives include the identiﬁcation of the leading factors of soil function variety and to determine the spatial heterogeneity of the soil function. A detailed topographic survey was carried out on a key site (35 hectares), 157 soil, and 34 geobotanical descriptions were made. The main factor of soil and plant cover differentiation is the redistribution of soil moisture along the microrelief. Redistributed runoff value was modelled in SIMWE and used as a tool for spatial prediction of soils due to their role in a habitat for plant communities’ functional context. The main methods of the study are the multidimensional scaling and discriminant analysis. We model the composition of plant communities (accuracy is 95%) and Reference Soil Group (accuracy is 88%) due to different soil moisture conditions. There are two stable soil habitat types: mesophytic communities on the Phaeozems (with additional water runoff more than 80 mm) and xerophytic communities on Chernozems (additional runoff less than 55 mm). A transitional type corresponded to xero- mesophytic communities on the Phaeozems with 55–80 mm additional redistributed runoff value. With acceptable accuracy, the habitat for natural plant communities based on soil function model predicts the position of contrastingly different components of biota in relation to their soil moisture requirements within the virgin forest-steppe of the Central Russian Upland.

Keywords: Chernozems; Phaeozems; Central Chernozem Reserve; predictive soil mapping

1. Introduction The most important ecological function of the soil as a component of the environment is to maintain the balance and sustainability of the biosphere under the anthropogenic impact and climate change [1–3]. The direct methods for soil ecological functions evaluation are limited, that is the main cause why researchers focus on indirect indicators of the soil systems functioning: the soil properties—i.e., the chemical composition of the different soil phases [4–7], soil structure [4,8,9], etc.—and the properties of environment connected to the soil by material—energy interactions. The latter include the species composition and productivity of plant communities [10–12] and in a broader sense, serve as the basis for landscape indication when mapping hard-to-observe properties of a landscape from lightly observed factors. The soil is the habitat for the plants—the environment of root nutrition and a seed pool. The study of the soil–vegetation interactions characterizes the features of the habitat for plant soil functions [2,13,14]. Soil structure (aggregate composition and pore space), chemical composition

Soil Syst. 2019, 3, 19; doi:10.3390/soilsystems3010019 www.mdpi.com/journal/soilsystems Soil Syst. 2019, 3, 19 2 of 12

(including organic carbon and nutrients concentration), and moisture characteristics are the key properties for the plant’s growth [4,9–11,15–17]. The connection between soil properties and plant parametersSoil Syst. are2018, implemented 2, x FOR PEER REVIEW in the global and regional ecological scales [18–21]. 2 of 13 Our study site is located in the virgin forest-steppe of East European plain. The climatic conditions (including organic carbon and nutrients concentration), and moisture characteristics are the key of the forest-steppe with favorable water-air and thermal regime contribute to the formation of the properties for the plant’s growth [4,9–11,15–17]. The connection between soil properties and plant mostparameters fertile soils are among implemented all the naturalin the global land and areas regional [22]. However,ecological scales the diversity [18–21]. of edaphic conditions determinesOur the study wide site variation is located of soil in the moisture virgin andforest-steppe aeration regimesof East European and, as a plain. result, The variety climatic in plant speciesconditions composition. of the forest-steppe We suppose with that favorable the soil moisturewater-air isand the thermal leading regime factor contribute controlling to the species compositionformation in of the the virgin most forest-steppe,fertile soils among as it all has the been natural qualitatively land areas shown[22]. However, [23]. the diversity of Theedaphic aim conditions of our work determines was the the digital wide variation mapping of ofsoil the moisture habitat and for aeration natural regimes plant communities’and, as a soil function,result, variety soil moisturein plant species conditions composition. and related We suppose plant speciesthat the compositionsoil moisture is in the virgin leading forest-steppe factor landscapescontrolling of V.V. species Alekhin composition Central in Chernozem the virgin forest State-steppe, Reserve as it (Kursk has been region, qualitatively Russia). shown The [23]. objectives The aim of our work was the digital mapping of the habitat for natural plant communities’ soil were to: (1) quantitatively characterize the relationship of the water regime and soil properties, function, soil moisture conditions and related plant species composition in virgin forest-steppe the compositionlandscapes of of V.V. plant Alekhin species; Cent (2)ral deﬁne Chernozem the different State Reserve habitat (Kursk types; region, (3) create Russia). and The verify objectives the habitat for a plantwere to: communities (1) quantitatively map. characterize We characterized the relation theship soil of forming the water factors regime features and soil forproperties, each section the of the regularcomposition grid withof plant resolution species; (2) 2.5 define m. The the redistributed different habitat runoff types; value (3) create was modelledand verify inthe SIMWE habitat [24]. Multidimensionalfor a plant communities scaling and map. discriminant We characterized analysis the were soil forming used to factors determine features the for relationship each section between of surfacethe runoff, regular soil grid moisture with resolution regime, 2.5 and m. The plant redi speciesstributed composition. runoff value The was main modelled hypothesis in SIMWE of our[24]. work is thatMultidimensional the plant species scaling composition and discriminant is the result analysis of a complexwere used interaction to determine in the the relationship sequence: between soil forming factorsurface (topography)—soil runoff, soil moisture forming regime, process and plant (redistribution species composition. of surface The runoff)—soil main hypothesis properties of our work (water and airis that regime)—soil the plant species function—species composition is the composition result of a complex of vegetation. interaction in the sequence: soil forming factor (topography)—soil forming process (redistribution of surface runoff)—soil properties (water 2. Materialsand air regime)—soil and Methods function—species composition of vegetation.

The2. Materials study was and completedMethods on the sub-horizontal part of the interﬂuve and gentle slope in the southwesternThe study part of was the completed Central Russian on the sub-horizontal upland of the pa Eastrt of European the interfluve Plain and to thegentle south slope of in the the Kursk city (Figuresouthwestern1) within part theof the Streletsky Central Ru steppessian upland cluster of of the the East V.V. European Alekhin Plain Central to the Chernozemsouth of the Kursk Reserve. ◦ The averagecity (Figure annual 1) within air temperature the Streletsky of steppe study clus areater isof 5.7the V.V.C, theAlekhin amount Central of precipitation Chernozem Reserve. is 590 mm. Soil-formingThe average parent annual materials air temperature are loess-like of study loams, area underlain is 5.7°C, the with amount 8 m byof chalkprecipitation deposits. is 590 mm. Soil-forming parent materials are loess-like loams, underlain with 8 m by chalk deposits.

FigureFigure 1. Study 1. Study area area (a) ( areaa) area location location (blue (blue square) square) relativelyrelatively natural natural areas areas and and large large cities, cities, (b) typical (b) typical topographytopography of central of central Russian Russian upland upland and and Central Central ChernozemsChernozems State State Reserve, Reserve, red red square—key square—key plot plot (coordinates(coordinates of the of center—51.5727N,the center—51.5727N, 36.0926E)). 36.0926E)).

Soil Syst. 2019, 3, 19 3 of 12

SoilKey Syst. plot 2018, area2, x FOR is PEER about REVIEW 35 ha. Its topography was studied in detail by GNSS surveys.3 of 13 Two Stonex S9III+ GNSS receiver were used. In fixed RTK regimes the positioning accuracy of horizontal coordinatesKey isplot 8 mm,area is vertical about 35 coordinates—15 ha. Its topography mm. was studied About in 10,000 detailof by points GNSS surveys. with accurately Two Stonex defined coordinatesS9III+ GNSS and height,receiver werewere usedused. for In digital fixed RTK elevation regimes model the (DEM)positioning creation. accuracy The of DEM horizontal resolution is coordinates is 8 mm, vertical coordinates—15 mm. About 10,000 of points with accurately defined 2.5 m, that is equal to distance between points. The ordinary kriging method [25] was used for DEM coordinates and height, were used for digital elevation model (DEM) creation. The DEM resolution creation.is 2.5 m, It was that completedis equal to distance in SAGA between GIS. Parameters points. The ordinary of calculation kriging are: method search [25] distance—100 was used for DEM m, max of pointscreation. in search It was distance—60, completed in variogramSAGA GIS. Parameters approximates of calculation by 4th degree are: search polynomial distance—100 function. m, max (Figure of 2b). Thepoints topography in search features distance—60, of the variogram study area approximates include two by 4th parts: degree the polynomial sub-horizontal function. interfluve (Figure with ◦ sinkholes2b). The and topography the northern features slope of the to 1.5study, complicated area include bytwo the parts: hollows the sub-horizontal of a complex interfluve genesis with described in [19sinkholes,26]. and the northern slope to 1.5°, complicated by the hollows of a complex genesis described in [19,26].

FigureFigure 2. ( a2.) ( Landa) Land use use types types (I—mowing (I—mowing steppe, II II—irgin—irgin steppe, steppe, III—pasture) III—pasture) and ( andb) DEM (b) DEMof key of key plot and soil survey points (yellow circles). plot and soil survey points (yellow circles). The surface runoff values were modelled by SIMWE in the GRASS GIS in the local neighborhood The surface runoff values were modelled by SIMWE in the GRASS GIS in the local neighborhood of each pixel. The model input parameters are the slope steepness, horizontal and vertical curvature, of eachthe pixel.amount The of excess model precipitation, input parameters infiltration are the rate, slope water steepness, diffusion horizontalcoefficient, surface and vertical roughness curvature, the amountcoefficient of [19]. excess Three precipitation, first parameters infiltration are calculating rate, waterin GIS by diffusion standard coefficient, methods [24]. surface Other roughness four coefficientparameters [19]. are Three equal first for the parameters study area are and calculating introduced ininto GIS the by model standard as constant. methods The [amount24]. Other of four parametersexcess precipitation are equal forand the infiltration study area rate and were introduced taken frominto regional the modelliterature as constant.[27]. WaterThe diffusion amount of excesscoefficient precipitation and surface and roughness infiltration coeffi ratecients were were taken compiled from from regional [28,29]. literature [27]. Water diffusion coefficientThe and soil surface survey points roughness are arranged coefficients for description were compiled of all the from diversity [28, 29of ].soils (Figure 2b). Totally, 157The boreholes soil survey with points a depth are of arranged 1 to 6 meters for description were completed. of all Detailed the diversity descriptions of soils of (Figure the horizons2b). Totally, 157 boreholeswere completed with according a depth of to 1 [30], to 6 the meters depth were of the completed. secondary Detailedcarbonates descriptions and the boundaries of the horizons of genetic horizons were determined; soil samples were taken from depth 0–10, 10–20, 30–40, 50–60, 70– were completed according to [30], the depth of the secondary carbonates and the boundaries of genetic 80, 90–100 cm. Air-dry soil samples were crushed and sifted through a sieve with a mesh size of 0.25 horizonsmm. In were samples, determined; the humus soil content samples was det wereermined taken by fromthe Tyurin depth method 0–10, [31]. 10–20, 30–40, 50–60, 70–80, 90–100 cm.Three Air-dry land-use soil types samples are presented were crushed on this andarea: sifted virgin, through mowing asteppes sieve withand pasture a mesh (Figure size of 2a). 0.25 mm. In samples,Detailed the humus geobotanical content descriptions was determined were perfor bymed the Tyurinat 34 sampling method plots: [31]. the plant communities andThree the land-useabundance types of species are presented were studied. on thisThe area:sampling virgin, plots mowing had size steppes 1 m by 1 and m. 19 pasture descriptions (Figure 2a). wereDetailed completed geobotanical in mowing descriptions steppe, 10 descriptions were performed in virgin at steppe, 34 sampling and 5 descriptions plots: the plant in pasture. communities and the abundance of species were studied. The sampling plots had size 1 m by 1 m. 19 descriptions were completed in mowing steppe, 10 descriptions in virgin steppe, and 5 descriptions in pasture. Soil Syst. 2019, 3, 19 4 of 12

All the raw materials: DEM, runoff modelled map, soil and vegetation descriptions are available by the link in supplementary materials. Species diversity has been determined on 1 m2 observational plots. Steppe meadows have a high level of species diversity per area (mean value for all releve is 23 species). This value is similar to Caucasus and Alps mountain meadows’ species richness evaluation and higher than mires species diversity in these mountain ranges [32]. The quantitative difference between relatively species richness 2 2 1 per area of 1 m and 100 m for meadows is around /2, that is higher compared with pine forests 1 1 2 ( /3) and beech forests ( /4) in Great Britain [33]. This allows one to consider the habitat’s area of 1 m as a basic to species and phytocoenosis ecological evaluation. The processing of the soil features and vegetation cover data, mapping and assessment of their accuracy was carried out in the SAGA GIS and programming language R. SAGA GIS was used for DEM creation and topography features calculation [34]. The maps evaluation was completed in R, packages ‘MASS’ (linear discriminant analysis) and ‘Ithir’ (accuracy assessment). For soil and vegetation cover prediction a relatively simple and widely used method Linear Discriminant Analysis (LDA) was used. LDA is a statistical method used to ﬁnd linear combinations of features that separate two or more classes of objects or events (discrete sizes) with the best accuracy [35,36]. Landolt ecological indicator values have been used for ecological evaluation of plant communities and habitats by species composition [20]. We have been taken rank species values for 8 ecological indicators (moisture, reaction, nutrient, humus, dispersion, light, temperature, continentality). Ecological indicator values have a wide application in ecological and geobotany investigation for determining the main gradients of species and plant communities’ relatively wide spectrum of factors [37–39]. Non-metric multidimensional scaling (NMS-ordination) [40] has been used for evaluation of plant communities’ variation according to complex of ecological factors with the application of Past program (v. 3.21) [41].

3. Results

3.1. Soils of the Study Site and the Factors Controlling Soil Properties Variety The soils of the study site represent two reference groups: Chernozems and Phaeozems. A common feature of all the studied soils is the formation of a very dark, well-structured, thick (over 50 cm, Pachic qualifier) and organic carbon-rich (more than 5%, Hyperhumic qualifier) chernic horizon in the upper part of the soil profile. The Chernozems have the calcic horizon or a layer with protocalcic properties. In Phaeozems, the secondary carbonates are absent. Some Phaeozems temporarily saturated with surface water for a period long enough that allows reducing conditions to occur (Stagnic qualifier) or having uncoated silt and sand grains on structural faces in the lower part of a humus horizon (Greyzemic qualifier). Thus, the pedodiversity in key plot is presented by 5 different soils: Haplic Chernozem Clayic Hyperhumic Pachic and Chernic Phaeozem Clayic Hyperhumic Pachic on the interfluve, Calcic Chernozem Clayic Hyperhumic Pachic on the convex hills, including relic marmot’s burrows, Greyzemic Chernic Phaeozem Clayic Hyperhumic Pachic and Stagnic Chernic Phaeozem Clayic Hyperhumic Pachic in hollows and closed depressions. The key properties for the soil biota are soil structure (aggregate composition and pore space), chemical composition (including organic carbon and nutrients concentration), moisture characteristics [4,9,15]. Chernozems and phaeozems are the most fertile soils all over the world [22], especially in virgin landscapes; the thick, well-structured, organic carbon-rich chernic horizon is formed in the upper parts of all the studied soils. The main difference between soils in the key site is the moisture conditions and the secondary carbonate presence or absence features. It was found, that average water content in chernozems is 280 mm, in phaeozems—340 mm. The water content in 1 m layer of soils was measured at the beginning of growing season. The latter is directly related to the soil Soil Syst. 2018, 2, x FOR PEER REVIEW 5 of 13

Soil3.2. Syst. Vegetation2019, 3, 19 Cover of Key Area and Relation of Vegetation Characteristics with Soil Moisture. 5 of 12 Vegetation cover of key areas is presented by plant communities, which are typical for interfluve waterhabitats regime, in the as forest-steppe it has been shownzone. Steppe in field meadow researchs and and herb modeling grass steppes [42,43 (sparse], including sod mesophytic in areas adjacent tograsses the key with site the [23 ].dominance Therefore, of weArrhenatherum assume that elatius the, differenceBromopsis riparia in soil, less moisture often Dactylis is the mainglomerata factor that and Calamagrostis epigeios) are presented. Dense sod xerophytic grasses species (Stipa pennata and influences on the soil habitat conditions variety at the key site. Festuca valesiaca) are co-dominante in some cases. For all phytocoenoses meadows-steppe herb 3.2.species Vegetation have Covera high oflevel Key of Area activity and Relationaccording of to Vegetation their frequency Characteristics and abundance. with Soil Meadow-steppe Moisture meso-xerophytic taproot (Onobrychis arenaria, Potentilla argentea, Viola rupestris) and mesophytic short rootedVegetation (Stachys cover officinalis, of key Centaurea areas is presentedjacea) perennial by plant species communities, are most frequent. which are Annual typical species for interfluve habitats(Melampyrum in the argyrocomum, forest-steppe Rhinanthus zone. Steppe minor meadows) are rare. and herb grass steppes (sparse sod mesophytic grassesThe with ordination the dominance scheme ofpresentsArrhenatherum some vector elatiuss , ofBromopsis environmental riparia , lessfactors often calculatedDactylis of glomerata andcommunitiesCalamagrostis according epigeios to) Landolt are presented. ecological Dense indica sodtor values xerophytic (Figure grasses 3a) [20]. species Landolt (Stipa ecological pennata and indicator scales include values of 8 ecological parameters (moisture, reaction, nutrient, humus, Festuca valesiaca) are co-dominante in some cases. For all phytocoenoses meadows-steppe herb dispersion, light, temperature, continentality). The scales allow us to estimate the ecological features species have a high level of activity according to their frequency and abundance. Meadow-steppe of the communities in accordance with the parameters. The ecological status for each releve is meso-xerophyticdetermined by the taproot ratio of (theOnobrychis sum of environmental arenaria, Potentilla values for argentea, all species Viola included rupestris in releve) and to mesophytic the shorttotal rooted number (Stachys of species officinalis, [44]. Ecological Centaurea values jacea have) perennial a significant species connection are most with frequent. two axes Annualof NMS- species (Melampyrumordination. argyrocomum, Rhinanthus minor) are rare. TheThe ordination first axis describes scheme the presents differences some in vectors diversity of and environmental abundance of factors speciescalculated at sample plots of communities due accordingto differences to Landolt in landuse ecological types. The indicator observed values plots (Figure corresponded3a) [ 20 to]. Landoltdifferent ecologicalland use types—the indicator scales includemowing values steppe, of virgin 8 ecological steppe and parameters pasture are (moisture, forming to reaction,isolated areas nutrient, on the humus,ordination dispersion, scheme. light, temperature,The accuracy continentality). of land use types The allocation scales allowby this us axis to estimateis 88%. the ecological features of the communities The second axis describes the observed differences in diversity and abundance of species in in accordance with the parameters. The ecological status for each releve is determined by the ratio of relation with soil moistening. In the attribute space of the axis the studied points are forming two therelatively sum of environmentalseparate groups valueswith a boundary for all species value includedof the axis in 2: releve −0.025 to(Figure the total 3b). numberThe accuracy of species of [44]. Ecologicaldetermiation values of vegetation have a significant types in relation connection with the with value two of axes redistributed of NMS-ordination. runoff is 95%.

FigureFigure 3. 3.(a )( distributiona) distribution of vegetationof vegetation plots plots on Non-metricon Non-metric multidimensional multidimensional scaling scaling (NMS-ordination) (NMS- andordination) their correlation and their withcorrelation Landolt with ecological Landolt ecolog indicatorical indicator values. values. The similarity The similarity index index is the is the Euclidean 2 2 distance.Euclidean Coordinate distance. Coordinate 1: R = 0.374; 1: R² Coordinate= 0.374; Coordinate 2: R = 2: 0.1775. R² = 0.1775. A—moisture, A – moisture, B—reaction, B – reaction, C—nutrient, C – D—humus,nutrient, D – E—dispersion,humus, E – dispersion, F—light, F – light, G—temperature, G – temperature, H—continentality.H – continentality. Types Types of land of use: land use: 1—mowing;1 – mowing; 2—virgin; 2 –virgin; 3—pasture;3 – pasture; andand ((bb)) thethe allocatedallocated vegetation vegetation types types in in the the attribute attribute space space of of water runoff and Axis 2 of NMS-ordination: 1—xero-mesophilic communities; 2—mesophilic communities.

The ﬁrst axis describes the differences in diversity and abundance of species at sample plots due to differences in landuse types. The observed plots corresponded to different land use types—the mowing steppe, virgin steppe and pasture are forming to isolated areas on the ordination scheme. The accuracy of land use types allocation by this axis is 88%. Soil Syst. 2019, 3, 19 6 of 12

Soil Syst.The 2018 second, 2, x FOR axis PEER describes REVIEW the observed differences in diversity and abundance of species6 of in13 relation with soil moistening. In the attribute space of the axis the studied points are forming two water runoff and Axis 2 of NMS-ordination: 1 - xero-mesophilic communities; 2 – mesophilic relatively separate groups with a boundary value of the axis 2: −0.025 (Figure3b). The accuracy of communities. determiation of vegetation types in relation with the value of redistributed runoff is 95%. The common scheme of topography-soil-vegetation relationsrelations isis presentedpresented atat FigureFigure4 4.. In In concaveconcave topographytopography forms withwith additionaladditional waterwater inflowinflow leachingleaching waterwater regimeregime isis typical.typical. At the flatflat andand convexconvex reliefrelief forms,forms, thethe chernozemschernozems withwith periodicallyperiodically leachingleaching waterwater regimeregime areare forming.forming. TheThe Landolt moisture factor for for chernozems chernozems is is significantly significantly lower, lower, than than for for phaeozems. phaeozems. This This regularity regularity is iscomplement complement to tothe the difference difference of of 1meter 1meter layer layer soil soil water water content. content. Generally, Generally, in in concave formsforms withwith additional waterwater inflowinflow andand moremore moistenedmoistened soilssoils areare formingforming mesophiticmesophitic phytocenoses.phytocenoses.

Figure 4. The common scheme of topography-soil-vegetation relations. Figure 4. The common scheme of topography-soil-vegetation relations. For the ﬁrst vegetation type, shown at Figure4, the modeled runoff in less than 80 mm. Xero-mesophilicFor the first steppevegetation meadows type, shown are formed at Figure under 4, these the modeled conditions. runoff They inare less represented than 80 mm. by Xero- herb (mesophilicOnobrychis arenaria,steppe meadows Filipendula are vulgaris formed)—dense under sodthese grass conditions. (Stipa pennata, They Festucaare represented valesiaca)—sparse by herb sod(Onobrychis grass (Arrhenatherum arenaria, Filipendula elatius, vulgaris Bromopsis)—dense riparia )sod communities. grass (Stipa Among pennata, the Festuca herb species valesiaca xerophytic)—sparse speciessod grass (Onobrychis (Arrhenatherum arenaria, elatius, Linum Bromopsis perenne, Stachysriparia) communities. recta) prevail. Among the herb species xerophytic speciesPlots (Onobrychis with values arenaria, on the Linum second perenne, axis of NMS-ordinationStachys recta) prevail. less than -0.025 and with modeled runoff more thanPlots 80 with mm, values are characterized on the second by anaxis additional of NMS-ordination runoff more less than than 80 -0.025 mm. and Mesophilic with modeled herb (runoffFilipendula more vulgaris, than 80 Achillea mm, are millefolium, characterized Amoria by repens, an additional A. montana runoff)—sparse more sod than (Agrostis 80 mm. tenuis, Mesophilic Dactylis glomerataherb (Filipendula) and sedge vulgaris, (Carex Achillea michelii millefo)—sparselium, sod Amoria grass repens, (Arrhenatherum A. montana elatius,)—sparse Bromopsis sod (Agrostis riparia) steppetenuis, meadowsDactylis glomerata are developed) and sedge under ( theseCarex conditions.michelii)—sparse In these sod communities, grass (Arrhenatherum mesophytic elatius, herb andBromopsis grass speciesriparia) dominatesteppe meadows in plant are communities. developed under Among these the mesophyticconditions. herbIn these species communities,Amoria repens, mesophytic Stachys ofﬁcinalis,herb and Rumexgrass species acetosella dominateare characterized in plant co bymmunities. these communities. Among the Meso-xerophytic mesophytic herb species species have Amoria a low levelrepens, of Stachys coenotic officinalis, activity. Rumex acetosella are characterized by these communities. Meso-xerophytic speciesThe have runoff a low on relieflevel microformsof coenotic activity. determines the division of the herb-grass steppe meadows of the watershedThe runoff into two on groups.relief microforms With the constant determines dominance the division of mesophytic of the herb-grass grass species steppe (Arrhenatherum meadows of elatius,the watershed Bromopsis into riparia two) in communities,groups. With differences the cons aretant found dominance in the participation of mesophytic of dense grass sod grassesspecies ((StipaArrhenatherum pennata, Festuca elatius, valesiacaBromopsis). Theriparia composition) in communities, of grasses differences with a certainare found ratio in ofthe mesophytic participation and of dense sod grasses (Stipa pennata, Festuca valesiaca). The composition of grasses with a certain ratio of mesophytic and xerophytic species has also an important role. With soil moisture increasing the part

Soil Syst. 2019, 3, 19 7 of 12 xerophytic species has also an important role. With soil moisture increasing the part of dense grass and xero-mesophytic herb species is decreasing. Sparse sod grasses co-dominate in the condition of higher moisture habitats. The species of xero-mesophyte and mesophyte phytocenoses on Chernozems and Phaeozems are differed by their morphological traits (Table1). Mesophyte phytocenoses on the Phaeozems are characterized by an absence of short-rhizomatous steppe perennial species (Veronica spuria, Phlomis pungens, Serratula tinctoria, Peucedanum oreoselinum, Anthyllis macrocephala). These species grow on the Chernozems only. Long rhizomatous (Scorzonera purpurea, Sanguisorba ofﬁcinalis, Rumex acetosa, Polygala comosa) and stoloniferous (Tanacetum vulgare, Amoria repens, Veronica chamaedrys, Leonurus quinquelobatus, Euphorbia semivillosa) species are speciﬁc to mesophyte phytocenoses on the Phaeozems. These meadow-steppe, meadow and nemoral species are related to increased moisture conditions. As a rule, they grow with a low abundance, but they can also dominate (for example, Amoria repens participates in the herb (Fragaria viridis, Convolvilus arvensis, Achillea millefolium)—grass (Bromopsis riparia) meadows communities).

Table 1. Life forms plant diversity in communities on the chernozems and the phaeozems.

Soil Reference Groups Life Forms Chernozem Phaeozem Sod perennial species 7 7 Short rhizomatous perennial species 23 20 Long rhizomatous perennial species 19 23 Stoloniferous perennial species 26 29 Bulbous species 1 1 Annual species 5 6 Shrub and semi-shrub species 2 2 Summ 83 88

3.3. Prediction and Veriﬁcation of Habitat for Plants Soil Map Based on Soil Function In Sections 3.1 and 3.2, we substantiated the essential role of the soil moisture (or water regime of soils) on the formation of soils and vegetation types within the study area. The topography, that is a redistributor of precipitation, determines the water regime of the soils. Therefore, the map of the habitat for natural plant communities based on soil function of the sample plot area is based on the map of the redistributed runoff value in the local neighborhood of each pixel (Figure5). The spatial structure of two vegetation types (Figure6a) and reference soil groups (Figure6b) were modelled by LDA method from redistributed runoff values. The accuracy of vegetation types map is 95%, kappa = 0.86. The accuracy of soil map is 86%, kappa = 0.68. The change of the reference soil group from Chernozems to occurs when the value of the redistributed runoff value is 55 mm. It was found, that the chernozems soils are forming at runoff value less than 55 mm, phaeozems—at runoff value more than 55 mm. Change of vegetation type is typical at runoff value of 80 mm. At the third stage, soil and vegetation maps were combined into one by overlaying and 3 habitat for plant types were obtained. The area of their distribution is presented in Table2, and the spatial structure is shown in Figure7.

Table 2. Raster overlay cross-tabulation (at intersection—types of vegetation habitats (Figure7), in brackets—area in ha).

Soil Reference Groups chernozems phaeozems Vegetation types Xero-mesophytes 1(33) 0 Mesophytes 2(1.5) 3(1) Soil Syst. 2019, 3, 19 8 of 12 Soil Syst. 2018, 2, x FOR PEER REVIEW 8 of 13

Soil Syst. 2018, 2, x FOR PEER REVIEW 9 of 13 FigureFigure 5. 5. TheThe SIMWE SIMWE modelled modelled runoff runoff value, m; isohypses are drawn through 0.5 m.

The spatial structure of two vegetation types (Figure 6a) and reference soil groups (Figure 6b) were modelled by LDA method from redistributed runoff values. The accuracy of vegetation types map is 95%, kappa = 0.86. The accuracy of soil map is 86%, kappa = 0.68. The change of the reference soil group from Chernozems to occurs when the value of the redistributed runoff value is 55 mm. It was found, that the chernozems soils are forming at runoff value less than 55 mm, phaeozems—at runoff value more than 55 mm. Change of vegetation type is typical at runoff value of 80 mm.

FigureFigure 6.6. ((aa)) Vegetation Vegetation types types of of the the key key area: area: 1 - 1—xero-mesophilic xero-mesophilic communities; communities; 2 - mesophilic 2—mesophilic (composition(composition of of phytocenoses phytocenoses in inthe the text) text) and and (b) Soils (b) Soils reference reference groups groups map: 1 map: – chernozems; 1—chernozems; 2 - 2—phaeozem.phaeozem.

At the third stage, soil and vegetation maps were combined into one by overlaying and 3 habitat for plant types were obtained. The area of their distribution is presented in Table 2, and the spatial structure is shown in Figure 7.

Table 2. Raster overlay cross-tabulation (at intersection—types of vegetation habitats (Figure 7), in

brackets—area in ha).

Soil Reference Groups chernozems phaeozems Vegetation types Xero- mesophytes 1(33) 0 Mesophytes 2(1.5) 3(1)

Soil Syst. 2019, 3, 19 9 of 12 Soil Syst. 2018, 2, x FOR PEER REVIEW 10 of 13

Figure 7. SoilSoil habitat for plants types: 1 1—convex – convex topography fo formsrms with xero-mesophytes on the chernozems, 2 2—concave - concave topography forms forms with xero-mesophytes on on the phaeozems, 3 3—concave - concave topographytopography forms with mesophytes on the phaeozems.

4. Discussion Discussion Thus, we we can can indicate indicate the the boundary boundary values values of of thethe redistributedredistributed runoffrunoff value corresponding to a change in the functioning modes of the biogeochemical cycle in the series of increasing leakage of moisture: (1) (1) convex convex slopes slopes and and flattened flattened sections of interfluve interfluve with a zonal precipitation rate with additional moisture less than 55 mm; (2) hollows with additional atmospheric-leakage moisture less than 80 mm; (3) hollows and ravines with additional moistening more thanthan 8080 mm.mm. The resultsresults indicateindicate the the existence existence of aof stable a stable relationship relationship in the in topography-soil-vegetation the topography-soil-vegetation system. system.Topography Topography as a factor as of watera factor flows of redistributionwater flows re determinesdistribution extra determines moisture inextra the concavemoisture elements, in the concaveleaching elements, water regime leaching and water an increase regime in and moisture an incr reservesease in moisture in the phaeosems. reserves in Inthe such phaeosems. conditions In such(No. conditions 3 in Figure 7(№), vegetation3 in Figure is7), adapted vegetation to specific is adapted habitat to specific conditions, habitat in whichconditions, the abundance in which the of abundancemesophytic of species mesophytic increases. species In this increases. case, the In soil th perfomsis case, the as asoil water perfoms reservoir, as a accumulating water reservoir, it in accumulatingperiods of excessive it in periods moisture of excessive and giving mois upture in periods and giving of droughts. up in periods of droughts. The secondsecond stablestable area area is is formed formed by soilsby soils located located on flat on and flat convex and convex elements, elements, where non-leachingwhere non- leachingwater regime water and regime lower and moisture lower reservesmoisture prevail. reserves In pr suchevail. positions In such (No. positions 1 in Figure (No. 71), in the Figure abundance 7), the abundanceof mesophytic of mesophytic species decreases species and decreases xero-mesophytic and xero-mesophytic species predominant. species predominant. Medium zonezone (No.(No. 2 2 in in Figure Figure7) with 7) with phaeozem phaeozem where where periodically periodically excessive excessive moistening moistening defines defineshigher moisturehigher moisture reserves reserves relatively relatively to No. 3. to In No. this 3. case In this periodically-leaching case periodically-leaching water regime water is regime tipical isand tipical xero-mesophyte and xero-mesophyte phytocenoses phytocenoses are formed. are formed. Thus, habitathabitat soilsoil function function for for plant plant communities communities in relationships in relationships with soilwith function soil function is that through is that throughthe accumulation the accumulation of spring moisture,of spring adaptivemoisture, species adaptive in anspecies optimal in ratioan optimal and abundance ratio and are abundance growing, arethe growing, ecological the requirements ecological requirements of which are suitableof which for are each suitable local for habitat each conditions. local habitat conditions. On the whole, the mesophyte communites on Phaeozems includes more herbs and Carex species with bigger leaf area than the species of communities on Chernozem. In accordance with [12,45–48]

Soil Syst. 2019, 3, 19 10 of 12

On the whole, the mesophyte communites on Phaeozems includes more herbs and Carex species with bigger leaf area than the species of communities on Chernozem. In accordance with [12,45–48] the plants of mesophyte phytocenoses could increase hydraulic roughness, sediment retention, and minimize soil erosion in comparison with the plants of xero-mesophyte phytocenoses.

5. Conclusions This paper is devoted to the methodological approach linking the regional climatic data and plant species composition due to topography and soil features. Three primary conclusions can be drawn from our research:

1. Soil moisture is the leading factor that determines the soil habitat for natural plant communities’ conditions variety in the virgin forest-steppe at a local level; 2. There are two stable soil habitats for natural plant communities types: mesophytic communities on the phaeozems (with additional runoff more than 80 mm) and xerophytic communities on chernozems (additional runoff less than 55 mm). 3. Map of habitat for plant communities in relationships with soil function allows us to predict the species composition, its distribution and abundance, the ecological and coenotic structure of steppe meadows and plant communities’ ecological speciﬁcity in relation to their requirements for soil moisture conditions.

Our methodological approach allows mapping different soil habitat for plants areas and thus can be used in agro-landscape engineering: crop species selection for planting and herbaceous hedges designing due to their moisture needs. This digital approach of soil mapping may be useful at provisional and regulating ecosystem services assessment [49]. Future works should include these relationships between crops, landforms and soil features in precision agriculture management in forest-steppe to maximize the biological production and reduce environmental risk.

Supplementary Materials: The following materials are available online at https://yadi.sk/d/yi8q5nIwE85gbw. Author Contributions: Conceptualization—D.K., M.S. and N.L.; Methodology—N.L. and M.B.; Investigation—N.L. and D.K.; Analysis—N.L. and M.B.; Visualization—N.L. and M. Bocharnikov; Writing—Original Draft Preparation—N.L., M.S. and M.B. Funding: This research was funded by Russian Foundation for Basic Research, grant number 17-04-02217. Acknowledgments: The authors thank Nina Gricay, student of Lomonosov MSU, for geobotanical descriptions performance. The authors acknowledge anonymous reviewers for their comments and suggestions to improve this paper. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Bouma, J. Soil science contributions towards Sustainable Development Goals and their implementation: Linking soil functions with ecosystem services. J. Plant Nutr. Soil Sci. 2014.[CrossRef] 2. Greiner, L.; Keller, A.; Grêt-Regamey, A.; Papritz, A. Soil function assessment: Review of methods for quantifying the contributions of soils to ecosystem services. Land Use Policy 2017.[CrossRef] 3. McBratney, A.; Field, D. Securing our soil. Soil Sci. Plant Nutr. 2015, 61, 587–591. [CrossRef] 4. Calzolari, C.; Ungaro, F.; Filippi, N.; Guermandi, M.; Malucelli, F.; Marchi, N.; Tarocco, P. A methodological framework to assess the multiple contributions of soils to ecosystem services delivery at regional scale. Geoderma 2016.[CrossRef] 5. Jiang, J.; Wang, Y.; Yu, M.; Li, K.; Shao, Y.; Yan, J. Science of the Total Environment Responses of soil buffering capacity to acid treatment in three typical subtropical forests. Sci. Total Environ. 2016, 563–564, 1068–1077. [CrossRef][PubMed] 6. Li, S.X.; Wang, Z.H.; Miao, Y.F.; Li, S.Q. Soil Organic Nitrogen and Its Contribution to Crop Production. J. Integr. Agric. 2014, 13, 2061–2080. [CrossRef] Soil Syst. 2019, 3, 19 11 of 12

7. Wang, Y.; Zhao, X.; Guo, Z.; Jia, Z.; Wang, S.; Ding, K. Soil & Tillage Research Response of soil microbes to a reduction in phosphorus fertilizer in rice—Wheat rotation paddy soils with varying soil P levels. Soil Tillage Res. 2018, 181, 127–135. [CrossRef] 8. Lehmann, A.; Stahr, K. The potential of soil functions and planner-oriented soil evaluation to achieve sustainable land use. J. Soils Sediments 2010, 10, 1092. [CrossRef] 9. Rabot, E.; Wiesmeier, M.; Schlüter, S.; Vogel, H.J. Soil structure as an indicator of soil functions: A review. Geoderma 2018.[CrossRef] 10. Bashir, H.; Ahmad, S.S. Ordination classification of relationship of vegetation and soil’s edaphic factors along the roadsides of Wahcantt, Pakistan. J. King Saud Univ. Sci. 2018.[CrossRef] 11. Liu, S.; Hou, X.; Yang, M.; Cheng, F.; Coxixo, A.; Wu, X. Factors driving the relationships between vegetation and soil properties in the Yellow River Delta, China. Catena 2018, 165, 279–285. [CrossRef] 12. Faucon, M.; Houben, D.; Lambers, H. Plant Functional Traits: Soil and Ecosystem Services. Trends Plant Sci. 2017, 22, 385–394. [CrossRef][PubMed] 13. Drobnik, T.; Greiner, L.; Keller, A.; Grêt-Regamey, A. Soil quality indicators—From soil functions to ecosystem services. Ecol. Indic. 2018.[CrossRef] 14. Haslmayr, H.P.; Geitner, C.; Sutor, G.; Knoll, A.; Baumgarten, A. Soil function evaluation in Austria—Development, concepts and examples. Geoderma 2016.[CrossRef] 15. Clapperton, M.J.; Chan, K.Y.; Larney, F.J. Managing the Soil Habitat for Enhanced Biological Fertility. In Soil Biological Fertility; Abbott, L.K., Murphy, D.V., Eds.; Springer: Dordrecht, The Netherlands, 2007. 16. Koptsik, S.V.; Koptsik, G.N.; Livantsova, S.Y.; Berezina, N.A.; Vakhrameeva, M.G. Analysis of the Relationship between Soil and Vegetation in Forest Biogeocenoses by the Principal Component Method. Russ. J. Ecol. 2003, 34, 37–45. [CrossRef] 17. Cachovanová, L.; Hájek, M.; Fajmonová, Z.; Marrs, R. Species Richness, Community Specialization and Soil-Vegetation Relationships of Managed Grasslands in a Geologically Heterogeneous Landscape. Folia Geobot. 2012, 47, 349–371. [CrossRef] 18. Dufrene, M.; Legendre, P. Species assemblages and indicator species: The need for a flexible asymmetrical approach. Ecol. Monogr. 1997, 67, 345–366. [CrossRef] 19. Eremenko, E.A.; Panin, A.V. Lozhbinnyj Mezorel’ef Vostochno-Evropejskoj Ravniny; MIROS: Moscow, Russia, 2010. (In Russian) 20. Landolt, E. Ökologische Zeigerwerts zur Sweizer Flora; Veröffentlichungen des Geobotanischen Institutes der Eidg. Tech. Hochschule: Zurich, Switzerland, 1977; pp. 1–208. 21. Ehrmann, J.; Ritz, K. Plant:soil interactions in temperate multi-cropping production systems. Plant Soil 2013, 376.[CrossRef] 22. Montanarella, L.; Badraoui, M.; Chude, V.; Costa, I.D.; Mamo, T.; Yemefack, M.; Aulang, M.; Yagi, K.; Hong, S.Y.; Vijarnsorn, P.; et al. Status of the World’s Soil Resources (SWIR)—Main Report; FAO: Room, Italy, 2015. 23. Fishman, M.I. Chernozemnye kompleksy i ih svyaz’ s rel’efom na Srednerusskoj vozvyshennosti. Pochvovedenie 1977, 5, 17–29. (In Russian) 24. Neteler, M.; Bowman, M.H.; Landa, M.; Metz, M. GRASS GIS: A multi-purpose open source GIS. Environ. Model. Softw. 2012, 3, 124–130. [CrossRef] 25. Oliver, M.A.; Webster, R. Basic Steps in Geostatistics: The Variogram and Kriging; Springer: Berlin, Germany. [CrossRef] 26. Velichko, A.A.; Morozova, T.D.; Nechaev, V.P.; Porozhnjakova, O.M. Paleokriogenez, Pochvennyj Pokrov i Zemledelie; Nauka: Moscow, Russia, 1996. (In Russian) 27. Agrofizicheskaya Harakteristika Pochv-stepnoj i Suhostepnoj Zon Evropejskoj Chasti; SSSR: Moscow, Russia, 1977. (In Russian) 28. Šustić,D.; Tadić,Z.; Tadić,L.; Kržak, T. Hydrologic and hydraulic analysis of less studied watershed. Energy 2008, 1, 1–10. 29. Koco, Š. Simulation of gully erosion using the SIMWE model and GIS. Landf. Anal. 2011, 17, 81–86. 30. IUSS Working Group WRB. World Reference Base for Soil Resources 2014, Update 2015 International Soil Classification System for Naming Soils and Creating Legends for Soil Maps; World Soil Resources Report No. 106; FAO: Rome, Italy, 2015. Soil Syst. 2019, 3, 19 12 of 12

31. Arinushkina, E.V. Guide on Soils Chemical Analysis; Moscow State University: Moscow, Russia, 1970; p. 488. (In Russian) 32. Onipchenko, V.G.; Semenova, G.V. Comparative analysis of the floristic richness of alpine communities in the Caucasus and the Central Alps. J. Veg. Sci. 1995, 6.[CrossRef] 33. Hopkins, B. The species-area relations of plant communities. J. Ecol. 1955, 43, 409–426. [CrossRef] 34. Conrad, O.; Bechtel, B.; Bock, M.; Dietrich, H.; Fischer, E.; Gerlitz, L.; Wehberg, J.; Wichmann, V.; Böhner, J. System for Automated Geoscientific Analyses (SAGA) v. 2.1.4. Geosci. Model Dev. 2015, 8.[CrossRef] 35. Bell, J.C.; Cunningham, R.L.; Havens, M.W. Soil Drainage Class Probability Mapping Using a Soil-Landscape Model. Soil Sci. Soc. Am. 1994, 58.[CrossRef] 36. Webster, R.; Burrough, P.A. Multiple discriminant analysis in soil survey. Eur. J. Soil Sci. 1974, 25, 120–134. [CrossRef] 37. Moser, B.; Büntgen, U.; Molinier, V.; Peter, M.; Sproll, L.; Stobbe, U.; Tegel, W.; Egli, S. Ecological indicators of Tuber aestivum habitats in temperate European beech forests. Fungal Ecol. 2017, 29, 59–66. [CrossRef] 38. Odland, A. Interpretation of altitudinal gradient in South central Norway based on vascular plants as environmental indicators. J. Ecol. Indic. 2009, 9, 409–421. [CrossRef] 39. Smirnov, V.E.; Khanina, L.G.; Bobrovsky, M.V. Validation of the Ecological-Coenotical groups of vascular plant species for European Russian forests on the basis of ecological indicator values, vegetation releves and statistical analysis. Bull. Mosk. Obs. Ispit. Prir. 2006, 111, 36–47. (In Russian) 40. Clarke, K.R. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 1993, 18, 117–143. [CrossRef] 41. Hammer, Ø.; Harper, D.A.T.; Ryan, P.D. PAST: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontol. Electron. 2001, 4, 1–9. 42. Sanderman, J. Can management induced changes in the carbonate system drive soil carbon sequestration? A review with particular focus on Australia. Agric. Ecosyst. Environ. 2012, 155, 70–77. [CrossRef] 43. Vereecken, H.A.; Schnepf, J.W.; Hopmans, M.; Javaux, D.; Or, T.; Roose, J.; Vanderborght, M.H.; Young, W.; Amelung, M.; Aitkenhead, S.D.; et al. Modeling Soil Processes: Review, Key Challenges, and New Perspectives. Vadose Zone J. 2016, 15, 1–57. [CrossRef] 44. Korolyuk, A.Y. Ecological preferences of plants on the south of Siberia. Bot. Issled. Sib. Kazakhstana 2006, 12, 3–38. (In Russian) 45. Kervroëdan, L.; Armand, R.; Saunier, M.; Ouvry, J.; Faucon, M. Plant functional trait e ff ects on runo ff to design herbaceous hedges for soil erosion control. Ecol. Eng. 2018, 118, 143–151. [CrossRef] 46. Berendse, F.; van Ruijven, J.; Jongejans, E.; Keesstra, S. Loss of plant species diversity reduces soil erosion resistance. Ecosystems 2015, 18, 881–888. [CrossRef] 47. Lambrechts, T.; François, S.; Lutts, S.; Muñoz-Carpena, R.; Bielders, C.L. Impact of plant growth and morphology and of sediment concentration on sediment retention efficiency of vegetative filter strips: Flume experiments and VFSMOD mod. J. Hydrol. 2014, 511, 800–810. [CrossRef] 48. Gyssels, G.; Poesen, J.; Bochet, E.; Li, Y. Impact of plant roots on the resistance of soils to erosion by water: A review. Prog. Phys. Geogr. 2005, 29, 189–217. [CrossRef] 49. Adhikari, K.; Hartemink, A.E. Linking soils to ecosystem services—A global review. Geoderma 2016, 262, 101–111. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).