Soil Catenas in Archeological Landscapes
F. N. Lisetskii
B e lg o r o d S ta te U n iversity, B elg o ro d , R u ssia
A b s tr a c t— S oil catenas cro ssin g various arch eological ob jects (ramparts, ancient settlem en ts, burial m ounds, etc.) with an ag e o f 3 6 0 - 1 8 0 0 years have been studied in the su b zo n es o f dark chestnut so ils and typical, ordi
nary, and southern ch ern ozem s in the East European Plain. It is found that topography-induced variations in the thickness o f zonal soil types are not alw ays distinct. O ften, the thickness o f so il profiles d o es not depend on the slo p e gradient. In this context, it is su p p osed that the rate o f the form ation o f hum us horizons on slo p es ex ceed s that on lev eled surfaces. A com parative an alysis o f the th ick n ess o f hum us horizons along so il catenas offers ex cellen t p o ssib ilities for the determ ination o f the rates o f erosion and p e d o g en esis. T h e m eth o d o lo g y o f quan
titative assessm en t o f the potential rate o f so il form ation on slo p es is su ggested .
FO R M U LA TIO N O F T H E PRO BLEM
In the past few decades, considerable progress has been achieved in the study o f soil evolution. At present, we can trace the H olocene history o f autom orphic (autonom ous) zonal soils in the R ussian Plain. H ow ever, in many places, the norm al developm ent o f soils is com plicated by erosion and redeposition o f sedi
ments. In these cases, a routine study o f the soil profile m orphology should be supplem ented with the study o f dynam ic processes in the landscape [23]. The catenary approach based on the analysis o f geochem ically co n
jugated soils seem s to be very prom ising for this pur
pose.
The concept o f soil catena im plies the integral anal
ysis of the soil profile and the soil cover (pedological landscape) [12]. Soil catena was defined by M ilne [40]
as a sequence o f soils that develop from sim ilar (as a rule) rocks, but in different topographic positions (sum mits, slopes, footslopes, etc.). The catenary approach is widely applied in soil science [30, 4 1 -4 3 ]. T he differ
entiation o f surface runoff along the slope dictates the differences in the rates of soil formation and, hence, the differences in the morphology o f soils com posing the cat
ena. Gradually, the concept o f soil catena has acquired a new meaning. At present, it is applied successfully not only to the description o f hydrological processes, but also to the study of the soil cover pattern and its spatial and temporal evolution [15]. Gennadiev [7] suggested the concept of spatial-tem poral m odels of pedogenesis. To develop these m odels, we should study not separate soils, but soil com binations com posing the chronose- quence and allocated to different elem ents o f local topography, different kinds o f parent rocks, and differ
ent vegetation associations, i.e., pedotopocatenas, ped- olithocom binations, and pedophytocom binations.
Ecologists and geobotanists consider soil changes on slopes in relation to the changes in the biota. From this point o f view, catenas m ay serve as the objects for
com bined analysis o f soil evolution, plant successions, and the dynam ics o f anim al population [26]. In this context, the concept o f polyclim ax developm ent can be applied not only to the soils, but also to vegetation. The zonal type o f vegetation can be considered a com bina
tion of several clim ax associations of plants differenti
ated by the elem ents o f m icrotopography, or a mosaic o f topographic, m icroclim atic, and edaphic vegetation com plexes. R egular changes in the character o f vegeta
tion along the slope gradient allow us to distinguish corresponding landscape structures that are called land
scape strips [24, 29], slope m icrozones [25], or parage- netic natural com plexes [3]. The latter may be inte
grated into landscape regions characterized by the sim ilarity o f changes in natural com plexes along slope gradients [23]. In order to reveal the effect of separate factors on the catenary structure o f landscapes (and soils), it is reasonable to study such catenas, in which ju st one o f the factors is a variable, w hereas all the rest are relatively stable. In m acrocatenas, the soils devel
oping at different topographic levels may have different ages [1, 2, 35]; this circum stance com plicates the inter
pretation o f the results. M eanw hile, the study o f soil catenas crossing various archeological objects (ancient settlem ents, ram parts, burial m ounds, etc.) with a quite definite tim e o f the beginning o f soil form ation (the zero m om ent) allow s us to reveal the effect o f geom or- phic conditions (in particular, the slope gradient) on soil developm ent and to trace the history o f soil evolu
tion and plant successions along such relatively mono- chronous catenas.
This aim of this study is to determ ine out the regulari
ties of morphological changes of soils along the topocate- nas of a definite age and to estimate the potential rate of soil formation on the basis of quantitative data on the thickness o f soil humus horizons along the slope.
O B JEC TS AND M ETH O D S
Six soil catenas allocated to archeological objects o f different ages (from 360 to 1800 years) w ere studied in the subzones o f dark chestnut soils and typical, ordi
nary, and southern chernozem s o f the E ast E uropean Plain. The studies were conducted in the territories o f the R ussian Federation, Ukraine, and M oldova. The zero m om ents o f soil form ation within these catenas were established on the basis of archeological and his
torical data.
(1) B elgorod oblast, Russia. The defense line (aba
tis) with a total length o f nearly 800 km was co n
structed along the southern border o f the R ussian State in 1635-1646. A part of this defense line stretching from B elgorod to B olkhovets and Karpov represents an artificial levee (ram part). The beginning (July 1, 1636) and the end (S eptem ber 15,1637) o f the construction of this ram part, as well as its original geom etric param e
ters (the length, ca. 13 km; the width in the top, about 6.4 m, and the height about 3.2 m), are known from histor
ical records [14, pp. 75-76]. The ram part was sur
rounded by a ditch. At present, a part o f this ram part (4154 m, or 37.8% o f the original length) is preserved not far from the B olkhovets settlem ent [31]. We studied the catena from the top of the ram part to the bottom o f the ditch surrounding it; the 20.3-m -long trench was excavated. The soils around the ram part (background surface soils) are represented by m oderately deep, low- hum us typical chernozem s.
(2) Vulkunesht district of M oldova and B olgrad dis
trict (Odessa oblast) of Ukraine. The so-called Trajan rampart (126 km) was constructed by Roman soldiers in the beginning of the second century AD to protect the new Roman province o f Dacia [11]. We studied a catena across this fam ous ram part on the right bank o f the B oFshoi Yalpug River. The background soils are represented by shallow -calcareous low-hum us ordinary chernozem s.
(3) B elgorod-D niester district, O dessa oblast, Ukraine. The construction of the Zm iev ram part found in this area dates back to the m iddle o f the fourth cen
tury AD as evident from the archeological artifacts bur
ied under the ram part and dating back to the ancient set
tlem ent that existed in place o f the ram part in the second and third centuries AD. A topographic survey conducted in 1870-1877 in this area showed that the ram part stretched from the right bank of the D niester Lim an to the settlem ent of Akerman (at present, Belgorod-on-Dni- ester) and had a length of about 12 km. We studied a part o f this ram part near the settlem ent o f Sadovoe. The background soils are represented by m icellar-calcare- ous low-hum us ordinary chernozem s.
(4) O vidiopol district, O dessa oblast, U kraine. The settlem ent o f N ikonii was constructed by B yzantine G reeks in this area. The territory of the settlem ent occu
pies about 3.8 ha. The settlem ent was abandoned in the fourth century AD, and the soil form ation started on the cultural layer. We studied the soil catena on a gentle slope of southw estern aspect. At present, this is an idle
land covered with natural grasses. A relatively short stage o f the agricultural developm ent o f this land began in 1910. T he background soils are represented by loam y low -hum us southern chernozem s.
(5) C rim ea, U kraine (29 km to the w est o f the city o f Kerch). The Uzunlar or Akkosov rampart (36 km) was constructed in this area. Data on the time of its construc
tion are rather contradictory. In some places, the rampart was reconstructed and new portions o f earth were added to it. However, ju dg in g from the analysis o f the thick
ness o f hum us profiles o f soils form ed within this ram part, we can definitely say that it was constructed in the second century AD [19]. The background soils are rep
resented by m oderately deep, low-hum us typical cher
nozems.
(6) B erezan district, N ikolaev oblast, Ukraine. In the southern part o f the peninsula that separates the Bere
zan and Sositsk lim ans, the defense constructions o f the ancient settlem ent (M ys I) were discovered. These con
structions represent a system o f a ditch, a rampart, one m ore ditch, and a stony wall inside. They were created by Rom ans in the beginning o f our era [4, p. 58]. Later, during the reign o f A ntoninus Pius (138-161 AD), w hen barbaric tribes destroyed several settlements around the B ug Lim an but failed to attack the settle
m ents around the B erezan Lim an being stopped by R om an troops, the defense lines were renovated and the ditches were deepened [4, p. 134]. Until the middle o f the third century AD, these lines protected the ancient settlem ent. Later, it was abandoned. The background soils in this region are represented by loam y-clay, dark chestnut soils with residual solonetzic properties.
Along with soil catenas crossing these archeological objects, we also studied several natural catenas: (a) the catena in the G lubokaya gulch (Kom intern district, O dessa oblast) with full-profile southern chernozem s (Fig. 1); (b) the catena along the southern slope with plow ed southern chernozem s in the same region (length, 600 m; gradient, 2.5°); (c) the catena in the O vidiopol district (eastern slope with a gradient of 2.4°
and length o f 200 m; southern chernozem s); and (d) the catena in the Ivanovskii district (O dessa oblast) dis
playing shallow ordinary chernozem s on the slopes o f w estern and eastern aspects with lengths o f 630 and 780 m and gradients o f 2° and 3°, respectively.
The main m ethod o f our study consisted o f a very detailed description o f soil trenches dug along the cat
enas in com bination with a m orphom etric analysis of local topography (data o f geodetic leveling were used for this purpose) and phytocenotic studies (the species com position o f plant associations and the phytomass (dry weight) density were determ ined). The color of soil horizons was described in the M unsell system {Munsell Soil Color Charts, 1975).
W hile analyzing the susceptibility o f separate soil profiles com posing the catena to erosional processes, we used averaged values o f the relief function F{L, J) substantiated in [37]. T his function describes the effect
L ISE TSK II
Plant association s; p h ytom ass, g /m 2 fescue,
366-199
fescue- agro- th y m e - herbs meadow grass, pyron, fescue, and
104 75 75 agro-
pyron, 87
agropyron-herbaceous, 121-129
H , m
Fig. l.T h e catena with full-Holocene southern chernozems; rangeland. Genetic horizons: (1) A, (2) A B ), (3) Bk, (4 ) depth of effer
vescence, and (5) the upper boundary o f soft calcareous nodules.
o f topography on soil loss tolerance. A dding the correc
tion coefficients for the effect o f the aspect and shape o f a slope on soil loss tolerance [33], we obtain
F(L- J) = X rjJ"L!0.5 (1) where X r! is the coefficient depending on the aspect and shape o f a slope; J is the slope gradient, %; L is the slope length, m; and n is the exponent depending on the character o f the soil surface.
RESULTS
The study o f vegetation patterns along the catenas in com bination with data on plant successions on cultural layers o f different ages (from 1600 to 3500 years) [22]
proved that the vegetation is m ore responsive to the changes in the environm ental factors than the soils.
Thus, the above ground phytom ass o f plants grow ing over dated surfaces with an age o f 2 0 0 0 -3 0 0 0 years reaches 8 0-8 4% of the above ground phytom ass of full-H olocene plant associations. A t the sam e time, the thickness o f the hum us profile o f soils w ith an age o f 1600-1800 years reaches ju st 66% o f the thickness of full-H olocene soils; the thickness o f soils with an age of 2 2 0 0 -3 20 0 years com prises about 79% o f the thick
ness of full-H olocene soils (at least, in conditions o f the steppe zone).
As was found earlier [20], m ean annual rates o f the developm ent o f the hum us horizon o f soils during the Subatlantic period (the last 2500 years) do not exceed 0.2 0 -0 .2 5 mm for ordinary chernozem s, 0 .1 8 -0 .2 3 mm for southern chernozem s, and 0 .1 5 -0 .2 0 m m for dark chestnut soils. However, the decrease in the rate o f the developm ent o f soil hum us horizons from the north (the forest-steppe and typical steppe subzones) to the south (the dry steppe subzone) is especially typical for lev
eled autom orphic positions in soil catenas, i.e., for flat tops. The thickness o f the hum us profile on slopes is dictated not only by the rate o f hum us form ation, but also by the processes o f erosion and redeposition of sedim ents differentiated by the elem ents of m eso- and m icrotopography.
T he average thickness o f the hum us horizon in the sloping part o f the catena across the B elgorod ram part (length, 14 m; inclination, 9.8°) com prises 16.85 cm , w hich is approxim ately equal to the thickness o f the hum us horizon on the leveled top o f the ram part (H = 16.98 ± 1.54 cm ; C v = 5.1% ; data from 32 separate m easurem ents).
T he m orphology o f so ils form ed on the ramparts
Location Intensity o f erosion T hickness o f horizons, cm AH,
(or accu m u lation ) A A + A B m m /year
T he B o lk h o v ets site o f B elg o ro d rampart, typical chern ozem Sum m it (4 .6 -m -w id e) o f the rampart near the w ood en tow er
Upper portion o f the slo p e (1 5 .5 °)
L ow er (co n ca v e) part o f the slo p e (1 1 .3 °) and the fo o tslo p e N o n e W eak
W eak accum ulation
13
N on e Z m iev rampart (m icellar-calcareou s ordinary ch ern ozem ) L ev eled sum m it
Upper portion o f the eastern slop e Upper portion o f a g en tle northern slop e
W eak W eak M oderate
19 15 N o n e L ow er Trajan rampart (calcareou s ch ern ozem ) U pper portion o f the southeastern slo p e (3 °)
Upper portion o f the northeastern slop e (3°) Upper portion o f the northern slo p e (2 ° - 3 ° ) L ow er portion o f the northern slop e
L ow er co n ca v e portion o f the northern slo p e ( 3 ° - 5 ° ) A ditch along the rampart
W eak W eak W eak M oderate
W eak accum ulation A ccu m u lation
19
6
13 30 192 U zunlar rampart (ch ern ozem on com p act c la y e y sed im en ts) Sum m it
Upper lev eled part o f a slope
N o n e W eak
16 15 T he rampart en circlin g the M y s settlem en t (dark chestnut so il)
17 14 9
37 35 23
4 0 30 30 2 6 52 295
31 28
0.4 7 0 .4 0 0.53
0.22 0.21 0.1 4
0.20 0.1 5 0.1 5 0.13 0.2 6 1.48
0.1 7 0 .1 6
Sum m it N o n e 15 30 0.1 6
U pper lev eled part o f a slop e W eak 14 27 0.15
Upper part o f the southern slo p e (2 °) W eak 12 2 6 0 .1 4
L ow er part o f the southern slo p e (4°) W eak accum ulation 2 6 33 0.18
B ottom o f the ditch betw een the rampart and the w all A ccu m u lation 36 84 0 .4 6
In the catena allocated to the N ikonii settlem ent (1600 years), the sloping part (the length is about 22 m, and the average inclination is 6.1°) is characterized by som ewhat lower reserves o f the green phytom ass of herbaceous vegetation with the predom inance o f fescue grass in com parison with the upper leveled part o f this catena (according to the m easurem ents in May, the d if
ference is alm ost 2.2-fold). At the sam e tim e, the aver
age thickness o f hum us horizons on the slope o f this catena is ju st 3 cm low er than in its upper leveled part;
taking into account the hum us-rich B 1 horizon, this d if
ference increases to 7 cm . In the upper leveled part of the catena, the thickness o f the A and A + A B 1 layers reaches 39 and 66 cm, respectively (Fig. 2). Thus, within the past 16 centuries, the rates of formation o f humus pro
files of the soils on slopes and on leveled surfaces in the subzone o f southern chernozems have been approxi
mately similar (0.23-0.24 mm/year). The full-Holocene analogue o f this catena was studied on the slope o f north- northwestern aspect with a length o f 32 m and inclination of 5.55°. At present, this area is used for grazing. The soils along the slope are characterized by distinct variations
in the thickness o f hum us horizon. In the upper leveled part o f the catena, the slightly solonetzic southern cher
nozem has a 35-cm -deep A horizon; the thickness of the full humus profile (A + AB 1) reaches 55 cm. In the slop
ing part o f the catena (length, 22 m; inclination, 4.7°), the thickness o f the humus horizon decreases from 42 to 11 cm and the depth o f the upper boundary of the horizon of secondary carbonate accumulations decreases from 71 to 31 cm. Tliese changes in the morphology o f soil profiles are associated with the development o f erosion. The green phytomass o f plants covering the eroded part o f the slope is 2 .7 -4 .9 tim es low er than that on the leveled surfaces.
A n indistinct dependence betw een the total thickness o f hum us horizons and the relief function values calcu
lated according to equation (1) is observed. However, as well as for the soils o f the catena at the settlem ent of N ikonii, this dependence becom es m uch m ore distinct, if we consider not the total thickness o f the humus layer, but the ratio o f the thickness o f the transitional (A B ) horizon to the thickness o f the upper hum us hori
zon A (AB : A). It is interesting to note that the soils with sim ilar ratios o f AB to A are allocated to more
L ISE TSK II F, g /m 2
- 6 0 0 - 5 0 0 - 4 0 0 - 3 0 0
{200 T0
0 2 4 6 8 10 12 14 16 1 8 2 0 2 2 2 4 2 6 2 8 3 0 32
L , m
Fig. 2. The catena with southern chernozems at the Nikonii set
tlement (the fourth century AD). Designations: (1) A horizon, (2) AB1 horizon, (3) ceramic artifacts, and (4) limestone debris; F, above ground phytomass.
erodible sites in the full-H olocene catena as com pared to the 1600-year-old catena.
The peculiarities o f soil formation on the Zmiev ram part are conditioned by the inverse character o f stratifica
tion of sediments composing the rampart. Its upper layers were constructed from carbonate-rich horizons of form er soils. At present, in spite o f the 16-century-old history of carbonate leaching, the topsoil horizons o f the ram part are still richer in carbonates than the underlying horizons (5.6 versus 4.1% at the depth o f 37-161 cm, respectively). This circum stance im pedes the form ation of the hum us horizon in the soils o f the ram part. The hum us horizon (A) is underlain by a layer with distinct pedogenic structure, but w ithout hum ic substances. The thickness o f this layer form ed by physical and biogenic processes o f structuring varies from 5 to 7 cm. A sim ilar picture is observed in the soils o f the U zunlar ram part, where the upper horizons contain up to 11.8% of C a C 0 3.
The trenches dug across the defense lines o f the M ys settlem ent (catena 1 crossing the outer ditch and the ram part and catena 2 crossing the inner ditch and the stone wall; Fig. 3) display evident distinctions betw een the soils form ed on these artificial constructions within the past 1800 years. In several places, these newly form ed so ils w ere distu rb ed d u rin g the p lan tin g o f tw o fo rest strip s in the end o f the 1960s. T h e above g ro u n d p h y to m ass in the u p p er parts o f slopes an d on lev eled su rfa ces reaches 2 6 0 -5 0 0 g/m 2 (the determ ina
tions w ere m ade in June); it decreases on slopes and increases considerably in depressions (form er ditches), w here it m ay be as high as 1127 g/m 2 (the ditch in cat
ena 1).
T he study o f soil horizons on slopes, especially, the m easurem ent o f horizon thickness, should be m ade w ith due respect to the inclination o f the surface [13];
in other w ords, all the m easurem ents should be done along the norm al to the surface. In the catenas studied, the soils o f footslopes above the ditches have a m or
phology sim ilar to the m orphology o f background dark chestnut soils. They are com posed o f the up per A hori
zon, transitional B 1 horizon, and carbonate-rich B2k hori
zon; the relative proportion between these horizons (with respect to their thickness) is 1 : 0.3 : 0.6. At the transition to the ditch, where the slope gradient is minimal (l° -2 ° ), the proportion between these horizons changes in favor of the transitional B1 horizon (1 : 1.2 : 0.5).
The full-Holocene dark chestnut soil was studied 8 km to the north o f the M ys settlem ent. T he soil develops under the feather grass association and is classified as a loam y clay dark chestnut soil with residual solonetzic properties. It is com posed o f the follow ing sequence of horizons: AO (0 -8 cm ), the sod layer of dark brown (10Y R 4/4) color, with very fine granular structure; A (8-31 cm ), the hum us horizon o f brown color (10Y R 4/3), with loose granular structure, coprolites, and evi
dent traces of eluvial processes (siliceous pow dering on ped surfaces); B1 (3 1 -5 0 cm ), the transitional horizon o f brow nish gray color (10 YR 5/3) and blocky struc
ture; B 2k (5 0 -6 3 cm ), the carbonate-rich horizon (the effervescence is observed from 53 cm ) partly im preg
nated by hum ic substances, with a bright brown color o f the m ain m ass (7.5Y R 5/6) and a blocky-prism atic structure; and B C k (6 3-88 cm ), the carbonate-illuvial horizon, yellow ish brow n (10YR 5/6), prism atic, very com pact; soft calcareous nodules (the white eyes) appear at the depth o f 82 cm; gypsum crystals are found at the depth o f 2.7 m.
K now ing the absolute age o f soils on artificial geo- m orphic surfaces, we can estim ate the effect o f topo
graphic conditions o f the differentiation o f the soil cover along the catenas.
T he structural differentiation o f the soils at the foot
slopes o f catenas 1 and 2 (the M ys settlement) is very close to the structural differentiation o f full-Holocene dark chestnut soils, in which the relative proportion between the A, B 1, and B2 horizons is expressed as 1 : 1 :0.5. The
Fig. 3. The catenas with dark chestnut soils formed on defense lines o f the Mys settlement (the second century AD). Designations:
( /) A horizon, (2) B 1 horizon, (J) B2 horizon, (4) earthy rampart, (5) stone (limestone) wall, (6) the upper boundary o f the horizon with soft calcareous nodules, (7) ceramic artifacts and bones, and (8) inclusions of limestone debris.
soil profiles on the tops o f the ram part and the stone wall represent unique objects that allow us to judge about the rate of zonal soil-forming processes in autonom ous con
ditions during the Subatlantic period. According to our observations, the average rate of the development o f soil humus horizons in these conditions reaches 0.16 m m /year (or 1.9 t/ha per year). It is interesting to note that the rates o f hum us form ation on two different substrates (the loesslike loam o f the ram part and lim estone o f the stone wall) are approxim ately similar. In transitional positions (on slopes), the thickness o f the hum us layer (A + B l) averages 28 cm, or approxim ately 50% o f the average thickness o f this layer in the surrounding full- H olocene soils (the latter was calculated from data on 144 pits). Thus, the annual increase in the thickness of the hum us layer o f soils on slopes is approxim ately equal to that on leveled surfaces (0.15 m m /year). H ow ever, taking into account the erosional processes oper
ating on slopes, we can assum e that the actual rate o f hum us formation in these conditions is even higher than on leveled surfaces. It is interesting to note that the thickness of the hum us layer (A + B l) in background dark chestnut soils varies irregularly on slopes o f differ
ent gradients [18] with a general tendency tow ard an increase with increasing gradients. Thus, on leveled interfluves, it averages 40.5 cm (C v = 14.8%); on slopes with inclination o f 4 °-9 °, 45.7 cm (C v = 21.7% ); and on slopes o f 9 °-1 5 °, 49.5 cm (C v = 14.0%). D iscussing
this phenom enon, Larionov suggested that the reason for it lies in a w eak differentiation o f soil m oistening on slopes o f different gradients in conditions o f the semi- arid clim ate o f dry steppes. From my point o f view, the tendency tow ard an increase in the thickness o f humus horizons on slopes, w hich is seen both in the full- H olocene soils and in the soils that form ed on artificial geom orphic surfaces, attests to the m ore active pro
cesses o f hum us form ation on slopes [19].
The hum us horizon o f soils in the ditches has been form ed not only due to in situ processes, but also due to the input o f hum us-rich m aterial from adjacent slopes.
The rate o f its form ation is assessed at 0.4 mm/year. In the inner ditch (catena 1), the am ount o f accum ulated hum us-rich sedim ents reaches 2.7 t per m eter o f the ditch boundary. The ditch is surrounded by slopes of the outer ram part (from the north) and the inner stone wall (from the south), which served as the sources of the m aterial accum ulated in the ditch. The sedim ent transport along the ditch from the east to the west is also possible. The southern slope o f the rampart has the length of 10 m and the inclination o f 12° (catena 1, Fig. 3). It is know n that southern slopes are more susceptible to ero
sion than the northern ones. At the same time, the biolog
ical turnover o f substances on southern slopes is more active, which results in the more wide ratio of the active humus to the passive humus [8]. This allows us to assume
LISE TSK II that soil-forming processes (and, in particular, humus- forming processes) are more active on southern slopes than on northern ones. Taking into account the amount o f material eroded from the southern slope o f the rampart and accumulated in the ditch (1.3-1.5 t/ha per 1800 years), we can calculate the rate o f the developm ent o f hum us horizons on this slope, which equals 0.23 m m /year, or 2.4 t/ha per year. This is 1.26-1.44 tim es higher than the rate o f soil form ation in leveled autonom ous posi
tions in this area. S im ilar data were obtained from the analysis o f the m odels o f hum us form ation on slopes and in autonom ous positions [19].
D ISC U SSIO N
Starting from the w orks o f D okuchaev, it is believed that the m orphology o f soils on slopes can be ade
quately interpreted in term s o f the activity of erosional processes differentiated by the elem ents o f slope topog
raphy. This approach form s the basis for the diagnostics and classification o f eroded soils. In turn, it is based on the assum ption that the soils o f slopes w ere sim ilar to the soils o f leveled interfluves with respect to the thick
ness o f their hum us profiles before the beginning o f the agricultural developm ent, i.e., before the beginning o f active soil erosion. However, this assum ption is not always true [8, 9, 38].
Already, the first works o f soil scientists pointed to the fact that slope conditions (gradient, aspect, and shape) not only control the intensity of erosional processes but also create specific habitats for plants differentiated by the elements of slope topography. Thus, Kostychev [17]
argued that the differences in the humus content between the soils of different elements of slopes may be condi
tioned not by the effect o f erosion, but by the differentia
tion of the whole set o f soil-form ing factors, including the vegetation. A ccording to his data, the chernozem s in the forest-steppe zone contained up to 8.5% o f humus on leveled surfaces; 8.4% , on the slopes o f northern aspect; 8.2% , on the slopes o f eastern and western aspects, and 5.7% , on the slopes o f southern aspect. The study o f chernozem s in the P o d o l’skaya gubem iya perform ed by N abokikh in the beginning of the 20th century (1150 sam ples w ere analyzed) proved that the relative differences in the hum us content of chernozem s on slopes and on leveled surfaces did not exceed 11-12% .
A nalyzing the dependence o f the thickness o f hum us horizon on the character o f surface topography (slope gradient and aspect) for different types o f soils, Larionov [18] suggested that two groups o f soils should be distinguished: (a) soils in which the thickness o f the humus horizon is virtually independent from the relief conditions (provided that these soils have not been plowed) and (b) soils in which the thickness o f hum us horizons varies considerably by the elem ents o f surface topography. The first group includes chestnut soils, sierozem s, and m ountainous brown forest soils. The second group includes various subtypes o f chernozem s.
In this context, the use o f the param eter o f the total thickness o f hum us horizons in chernozem s seem s to be inappropriate for the analysis o f the rates o f ero
sional processes and soil-form ing processes on slopes.
T he study o f recent soils form ing on tailings in the course o f th eir biological rem ediation proves that these soils have the sequence o f m orphological horizons resem bling that in zonal soils. Soil as a holistic self- developing system can be characterized by certain pro
portions betw een its param eters. If these proportions are relatively stable, we can take them as the basis for soil diagnostics and classification. Prasolov [28] noted that in full-profile developed chernozem s, the propor
tion betw een the A and the A + B horizons is close to 1.0. T he ratio o f the thickness o f the layer A + AB to the thickness o f the A horizon was suggested as a criterion fo r the assessm ent o f the degree o f erosion o f cher
nozem s [32]. O nishchenko [27] developed a m ethod fo r the reconstruction o f genetic horizons o f eroded soils on the basis of the analysis o f the proportions betw een the thickness o f different genetic horizons in noneroded soils. Indeed, the horizons o f natural soils (A, B l , and B2) often correlate with one another with respect to their thickness (at least, in chernozem s).
T he analysis o f available data on the dynam ics of the w ater regim e o f steppe soils and on the rates of hum us form ation in these soils proves that the develop
m ent o f steppe soils proceeds unevenly. In other words, the periods o f active developm ent o f the soil profile alternate with the periods o f relative stability. In this context, the analysis o f proportions betw een the hori
zons (A : AB : B) may be o f great interest, as these pro
portions m ight reflect the differences in conditions of soil form ation during a certain stage o f soil develop
m ent. In general, these proportions are relatively stable for zonal soil types form ing in autom orphic (leveled interfluves) positions [34]. In order to reveal the effect o f slope conditions on the character o f soil form ation, w e may analyze the changes in the proportions o f soil horizons at different elem ents o f the slope. In particu
lar, the ratio betw een the thickness o f the transitional horizon (AB1 or B l) and the thickness o f the hum us horizon (A) was analyzed by us. T he characteristic tim es o f the developm ent o f these horizons are differ
ent. A ccording to Kovda, the characteristic tim e o f hum us accum ulation in the transitional horizon is
1.5 tim es lon ger than the characteristic tim e o f hum us accum ulation in the upper hum us horizon [16]. Thus, the AB : A ratio can be interpreted as the index o f the age o f a soil. T he greater the age, the w ider the ratio.
(This conclusion follows from the assum ption that the developm ent o f the hum us profile in soils proceeds dow nw ards).
In the catenas studied in the southern steppe and dry steppe subzones (Figs. 2, 3), the ratio o f AB (or B) to A decreases with an increase in the value o f the relief function F(L, /). The alteration o f this function from 60 to 130 leads to a decrease in the AB : A ratio from 1.1
to 0.1. At the same tim e, the total thickness of the humus profile (A + AB) does not depend on the relief function. The dependence o f the AB : A ratio on slope conditions may be interpreted in the follow ing manner.
The higher potential erodibility o f the soils on slopes (in com parison with the soils o f leveled surfaces) induces the transform ation o f soil m orphology so that the thickness o f the transitional AB horizon decreases at the expense o f an increase in the thickness o f the top
m ost A horizon. However, the latter horizon m ay be subjected to active erosion. In this case, its thickness decreases and the AB : A ratio increases. Thus, this ratio may serve as an index of soil erosion. In general, the anal
ysis o f the proportions betw een different soil horizons on slopes may be very inform ative for the assessm ent of the rates o f soil form ation and soil erosion on slopes.
The notion o f self-regulation of soil systems assumes more active soil formation in conditions o f denudation.
This is one of the important mechanisms that ensures the stabilization o f the dynam ic natural system “tectonic m ovem ents-soil form ation-denudation and accum ula
tion o f sedim ents.” L et us consider the m echanism s o f self-regulation in detail. It can be assum ed that these m echanism s have a sim ilar nature in the soils o f differ
ent ages, including eroded and deflated rejuvenated soils. D uring the initial stages o f soil developm ent, soil- form ing processes are nonequilibrium and lead to the self-organization o f the soil profile. Actually, a sim ilar effect is observed in recently eroded soils that are also in the nonequilibrium state with the environm ent. In particular, the natural balance betw een the rates o f hum ification and m ineralization is disturbed in eroded soils. As a result, as soon as the surface is covered with plants, the coefficient o f hum ification in eroded soils exceeds that in noneroded m ature natural soils [6]. It was proven that the rate o f hum us accum ulation on hum usless rock is three to four tim es m ore active than that on hum us-rich substrate [36]. It is know n that the hum us content in the upper layers o f soil-form ing rocks beneath the hum us layer reaches 1.37% in strongly eroded chernozem s and ju st 0.86-0.91 % in slightly and moderately eroded chernozem s [39].
Eroded soils are characterized by the nonequilib
rium state and by a higher coefficient o f hum ification in com parison with equilibrium noneroded zonal soils.
However, the potential for m ore active hum us form a
tion in eroded soils can only be realized during the peri
ods o f attenuation o f erosional processes both in annual and longer cycles (the periods of geom orphic stability) [10]. We can use two different m ethods to estim ate the rate o f hum us form ation in eroded soils on slopes. First, we can calculate it from data on the soil age and hum us content (table) with due account for the am ount o f hum us that has been rem oved from a particular soil in the course o f erosion. Second, we can use a special model describing the potential rate o f soil form ation on slopes. The latter m ethod seem s to be very prom ising for the design o f erosion-control m easures.
Let us consider the m odel o f the potential rate of soil form ation on slopes in detail. First, we can describe the changes in the total thickness o f the hum us layer along the slopes using the follow ing equation:
H ( L) = 1 / L H ( l ) d l , (2) w here H(L) is the average thickness o f the hum us layer o f soils along the line from the local divide to the given point located at a distance L from the divide and H([) is the thickness o f the hum us layer at an arbitrary point on the slope located at a distance / from the divide.
The thickness o f the hum us layer (H( L), mm) for a soil catena with the soils o f sim ilar age can be calcu
lated according to the follow ing equation:
H{ L ) = 1/L
(3) x [ 1 0 .8 5 g (F / / F , ) ° 37e x p (0 .0 0 4 4 Q )(l - ke~X')]dl, where Ff is the factual vegetation productivity, t/ha per year; Fz is the zonal (clim ate-controlled) productivity o f vegetation, t/ha per year; Q denotes energy expenses on soil form ation, M J/m 2 per year, according to [5]; g is the correction coefficient for the effect o f soil texture; t is the duration o f soil form ation, years; and k and X are em piric param eters, o f which the values for the main types o f soils in the R ussian Plain have been published in [19].
The changes in energy expenses on soil formation during the H olocene have been specially studied for autom orphic (zonal) soils o f steppe ecosystem s [21].
The results of these studies allow us to calculate the val
ues o f Q(t) and F(t) for shorter periods. The redistribu
tion o f heat and w ater resources by the elem ents of slope topography can also be described. The net pri
m ary production o f phytocenoses and its redistribution along the slope can be assessed from data on the biom ass o f the above ground parts o f plants or on the mort- m ass stored in the soil.
If we co nsid er a catena com posed o f full-Holocene soils, equation (3) can be sim plified via grouping of constant values:
H ( L ) = A / L ( g ( l ) F ^ 37 (I) exp (0.0044 Q ( l ) ) d l , (4) where A is an em pirical coefficient equal to 3.7 for the chernozem s o f the forest-steppe zone, 3.8 for ordinary chernozems, 4.4 for southern chernozems and dark chest
nut soils, and 6.0 for chestnut soils. Calculations by equa
tion (4) can be m ade via application o f well-known m ethods o f num erical integration (the m ethod of tra
peze or the Sim pson m ethod). In order to obtain the val
ues o f H averaged for a certain area, we have to use double integration taking into account the variations in distribution o f soil-form ing factors across and along the slope.
The equation for calculation of the thickness o f the hum us layer (and, hence, the rate o f soil form ation) on the slopes under croplands has the same shape, but it
1
L ISE TSK II has to be supplem ented with experim ental or calculated data on the rate o f annual soil erosion.
E quation (4) allows us to calculate the potential thickness o f the hum us layer o f soils located at different positions along a slope. This calculation can be m ade both for the full-H olocene soils and for the soils o f a certain age. Knowing the potential thickness o f the hum us layer, we can com pare it w ith factual data and, thus, estim ate the effect o f erosional processes. The study of catenary conjugations o f soils with known zero-m om ent of soil form ation, including the study of soil catenas in the areas o f archeological excavations, may be very prom ising for this purpose. In this respect, the collaboration o f soil scientists and archeologists might be very beneficial for both sciences.
C O N C LU SIO N S
(1) T he soils form ed on d ated su rfaces ( 3 6 0 - 1800 years) in the forest-steppe and steppe zones of the R ussian Plain are characterized by a relatively uni
form thickness o f the hum us horizon both on leveled surfaces and in the erosion-susceptible parts o f slopes.
This phenom enon can be explained by a higher rate of soil form ation on slopes as com pared to leveled inter
fluves.
(2) In soil catenas under natural vegetation, the thickness of the hum us layer virtually does not depend on the value o f the relief function, w hich takes into account the shape, aspect, gradient, and length o f a slope; at the sam e time, the increase in the relief func
tion value is accom panied by the general thinning of the hum us layer and a rapid increase in the ratio o f the thickness o f the transitional AB (or B 1) horizon to the thickness o f the A horizon.
(3) The ratio o f AB (B l) to A horizons reflects the degree of soil developm ent. W ith an increase in the age of soils and corresponding decrease in the rate of hum us horizon developm ent, the AB : A ratio increases.
(4) Special catenary studies o f soils in com bination with m odeling the potential rate o f soil developm ent in different parts o f slopes seem to be very prom ising for the assessm ent o f the rates o f soil erosion and soil for
mation on slopes.
R EFE R EN C ES
1. A derikhin, P.G., G eography and T opography o f S o ils in Central C h em o zem ic R eg io n s, P o c h v o v e d e n ie i p r o b lem y sel'sk o g o k h o zy a istv a (S o il S c ie n c e and A g ricu l
tural P roblem s), Voronezh: V oronezh. G os. U niv., 1979, pp. 3 8 -5 4 .
2. B altyanskii, D .M ., S o ils o f Central M oldavia: R egulari
ties o f T heir G eographical L ocation and M anagem ent A sp ects, A b s tr a c t o f C and. Sci. (G e o e r .) D is s e r ta tio n , C hisinau, 1970.
3. B erezh n oi, A .V ., S lo p e P aragenetic N atural C om p lex, Its Structure and Study, P rik la d n y e a s p e k ty izu ch en iya sovrem en n ykh la n d sh a fto v (Practical A sp e c ts o f Stu d y
ing the M odern L andscapes), V oronezh: Voronezh. G os.
U niv., 1982, pp. 9 3 - 1 0 7 .
4 . B u isk ik h , S .B ., F o rtifik a tsiya O l ’v iisk o g o g o su d a rstv a (p e r v y e vek a n a sh ei e ry ) (Fortification o f the O l’vian State: T h e First C enturies A D ), K iev: N aukova D um ka,
1991.
5. V olobuev, V .R., V veden ie v en ergctiku p o c h v o o b ra z o - v a n iy a (Introduction to S oil-F orm in g E nergetics), M o s
cow : N auka, 1974.
6. Ganzhara, N.F. and Ganzhara, L .N ., Relationship between the R ates o f Erosion and Formation o f a Hum us Horizon in Eroded Soils, O tsenka i kartirovanie ero zio m o o p a sn yk h i d eftya tsio n n o o p a sn yk h z e m e l' (A ssessm en t and M ap
pin g o f L ands Subjected to E rosion and D eflation ), M o s
cow : M o sk . G os. U niv., 1973, pp. 1 2 0 -1 2 5 .
7. G ennadiev, A .N ., P o c h v y i vrem ya : m o d e li ra zv itiy a (S o ils and Tim e: M o d els o f D ev elo p m en t), M oscow : M o sk . G o s. U niv., 1990.
8. G eorgi, A .A . and Sakhle, M .Z ., D ev elo p m en t o f S o ils on G u lly S lo p e s in the L eft-B ank Forest Step p e o f U krai
nian S S R , Tr. K h a r ’k. S -k h . Inst. im. V.V. D o k u ch a eva , 1984, v o l. 2 9 9 , pp. 6 5 - 7 0 .
9. G linka, K .D ., P o c h v y (S o ils), M oscow : G osizdat, 1923.
10. Gerrard, A .J., S o ils a n d L andform s. An In teg ra tio n o f G e o m o rp h o lo g y a n d P e d o lo g y , London, 1981. Trans
lated under the title P o c h v y if o r m y rel ’efa. K o m p lek sn o e g e o m o rfo lo g o -p o c h v e n n o e issle d o v a n ie , Leningrad:
N edra, 1984.
11. D zigovsk ii, A .N . and L isetskii, F.N., Pedochronological A spects o f Dating the N izhnii Trajan Rampart, A rkh eologiya i etn ologiya Vostochnoi E vropy: m aterialy i is sle d o v a n iy a (A rch eo lo g y and E th n ology o f Eastern Europe: M aterials and S tu d ies), O dessa, 1997, pp. 2 2 9 - 2 3 8 .
12. D o lo to v , V .A ., P rofile-R egional Soil C oncept, Izv. A kad.
N auk SSSR, Ser. B iol., 1971, no. 1, pp. 1 2 2 -1 2 7 . 13. Zakharov, S .A ., S o m e D ebatable Problem s o f M ountain
S o il S c ien ce, P o c h v o v e d e n ie , 1948, no. 6, pp. 3 4 7 -3 5 6 . 14. Izv. K ursk, g u b ern sk o g o o -v a k ra e v e d e n iy a , 1927, no. 5,
pp. 7 5 - 7 6 .
15. Karavaeva, N .A . and T im ofeev, D .A ., Integration o f Soil S c ie n c e and G eom orp h ology, Izv. A kad. N auk SSSR, Ser.
G eo g r., 1985, no. 6, pp. 1 3 5 -1 3 8 .
16. K ovda, V .A ., B u grovsk ii, V.V., Kerzhentsev, A .S ., and Z elen sk aya, N .N ., T he M odel o f Organic M atter Trans
form ation in the S oil for Q uantitative Study o f S oil F u n ction in E co sy stem s, D okl. A kad. N auk SSSR, 1990, vol. 3 1 2 , no. 3, pp. 7 5 9 -7 6 2 .
17. K ostychev, P.A., P o c h v y ch ern o zem n o i o b la s ti R o ssii (S o ils o f the C h em o zem ic Z on e o f R u ssia), M oscow : S e l’k h o zg iz, 1949.
18. Larionov, G .A ., E ro ziya i d e fly a ts iy a p o c h v (E rosion and D eflation o f S o ils), M oscow : M osk. G os. U niv., 1993.
19. L isetsk ii, F.N ., L aw s o f the Form ation o f a H um us H ori
zon in the Zonal S o ils o f the R ussian Plain, A grokhim . P o c h v o v e d ., 1990, no. 5 3 , pp. 3 - 7 .
2 0. L isetsk ii, F.N ., Study o f S o il- and L andscape-Form ing P r o cesses in the R eg io n s o f A n tic States in the Northern C oast o f the B lack Sea, Isto ric h e sk a y a g eo g ra fiya : ten- d e n ts ii i p e r s p e k tiv y (H istorical G eography: T endencies
and P erspectives), St. Petersburg: R ussk. G eogr. O - v o , 1995, pp. 7 4 - 8 3 .
2 1. L isetsk ii, F.N ., S p atial-T em poral A sse ssm e n t o f Plant Production as a Soil-F orm in g Factor, P o c h v o v e d e n ie ,
1997, no. 9, pp. 1 9 5 5 -1 0 5 7 .
2 2. L isetsk ii, F.N ., A u to g en ic S u ccessio n o f Step p e V egeta
tion in Postantic L andscapes, E k o lo g iy a (S v erd lo v sk ), 1998, no. 4 , pp. 2 5 2 -2 5 5 .
2 3. M e to d ic h e sk ie u k a za n iya p o la n d sh a ftn ym is sle d o - va n iya m d ly a s e l ’sk okh ozyaistven n ykh ts e le i (M eth od o lo g ica l G u id elin es for L andscape Stu d ies for A g ricu l
tural U se s), S h veb s, G.I. and S h ish ch en k o, P.G., E ds., M o sco w : V ses. A kad. S e l’skokhoz. N auk, 1990.
2 4 . M iller, G.P., M ountain L andscape Structure, Vestn.
M osk. Univ., Ser. G eogr., 1968, no. 3, pp. 8 8 - 9 2 . 25. M il’kov, F .N ., L andscape M icrozon ality on S lo p es,
N auchn. Zap. Voronezh. O td. G eogr. O -v a . SSSR, 1974, pp. 3 - 9 .
26. M ordkovich, V.G., Shatokhina, N .G ., and Titlyanova, A .A ., S tep n ye k a ten y (Steppe C atenas), N ovosib irsk : N auka,
1985.
2 7. O nishchenko, S .K ., D eterm ination o f D eform ation D egree in the U pper Part o f S o il P rofile by the G en etic H orizon R econstruction M ethod, D okl. A kad. N auk SSSR, 1976, vol. 2 3 1 , no. 6, pp. 1 4 6 2 -1 4 6 4 .
2 8. P o c h v y SSSR (S o ils o f the S o v iet U n io n ), Prasolov, L.I., E d., M oscow : A kad. N auk S S S R , 1939.
2 9. Proka, V.E., M o rfo lo g ich esk a ya stru k tu ra la n d sh a fto v i zem leu stro itel’noe proektirovanie (M orphological Struc
ture o f Landscapes and Land M anagem ent D esign), C h i
sinau: Shtiintsa, 1976.
30. R assel, E., P o ch ven n ye u slo viya i ro s t ra s te n ii (S o il C on d ition s and Plant G row th), M oscow : Inostrannaya Literatura, 1955.
31. R yapolov, V .M ., B elgorod Streak (Southw estern Scarp):
Problem s o f Preservation and M u seification , M a te ria ly n au ch .-prakt. k o n f: “ Yug R o ssii v p ro s h lo m i n a sto y a - shchem : isto riya , ekonom ika, k u l’tu r a ” (M aterials o f Sci.-Pract. C onf. “Southern Russia in the Past and Present:
H istory, E con om y, and C ulture”), B elgorod: Belgor.
G os. U niv., 1998, pp. 2 2 -2 4 .
3 2 . Sorokina, N .P., Statistical E valuation o f E rosion in Typ
ical C hern ozem s at the Kursk Experim ental Station, P o c h v o v e d e n ie , 1966, no. 2, pp. 9 1 - 9 6 .
33. S p ra vo ch n ik p o po ch vo za sh ch itn o m u zem le d e liy u (Text
book on Soil-P rotection Farm ing), K iev: U rozhai, 1990, pp. 2 0 - 3 8 .
34. Stepanov, I.N ., T he C onstancy o f the R atio betw een the T h ick n esses o f H um us and U nderhum us H orizons in M o d em S o ils, D okl. A kad. N auk SSSR, 1981, vol. 261, no. 6, pp. 1 4 3 7 -1 4 3 9 .
3 5. U rsu, A .F., P o c h v en n o -ek o lo g ich esk o e m ikroraion iro- va n ie M o ld a v ii (S o il-E c o lo g ic a l M icrozonation o f M o l
davia), C hisinau: Shtiintsa, 1980.
36. Fokin, A .D ., Participation o f D ifferent C om pounds from Plant R em ain s in the Form ation and R enew al o f Soil H um ic S u b stances, P ro b lem y p o c h v o v e d e n iy a (Prob
lem s o f S o il S c ie n c e ), M oscow : N auka, 1978, pp. 6 0 -6 5 . 3 7 . S h v eb s, G .I., T eoretich eskie o sn o v y e ro zio ved en iy a
(T heoretical Principles o f E rosion S tu d ies), Kiev:
V y ssh a y a Sh k ola, 1981.
3 8. Shurikova, V.I. and Erm olenko, G.P., C lassification o f Eroded Gray Forest Soils, Erozxya p o ch v i nauchnye osn ovy b o r ’b y s nei (Soil Erosion and Scientific Principles o f Its Control), M oscow : Pochv. Inst. im. V.V. D okuchaeva,
1985, pp. 6 4 - 6 8 .
3 9. Y arovenko, A .F., C h ernozem s on D on b ass S lo p es and T heir Properties, Tr. K h a r ’kov. S -K h . Inst., 1974,
vol. 196, pp. 8 2 - 8 8 .
4 0 . M iln e, G ., S o m e S u ggested U nits o f C lassification and M apping, Particularly for East A frican S o ils, S o il Res.,
1935, vol. 4 , no. 3, pp. 1 8 3 -1 8 9 .
4 1 . R uhe, R. V., G eom orphic Surface and the Nature o f Soils, S o il S ci., 1956, vol. 82, pp. 4 4 1 ^ t5 5 .
4 2 . S im on son , R .W ., A M u ltip le-P rocess M o d el o f S oil G en
esis, Q u a te rn a ry Soils. G eo A b s tr a c ts , 1978, pp. 1 -2 5 . 4 3 . V reeken, W.J., S oil-L an d scap e C hronogram s for Pedo-
ch ron ological A n a ly sis, G eo d e rm a , 1984, vol. 34, no. 2, pp. 1 4 9 -1 6 4 .
