5

Silva Balcanica, 17(1)/2016

WATER RUNOFF AND ERODED SOIL FROM INTENSIVE PRECIPITATIONS ON DIFFERENT LAND USE TYPES

IN SOUTH-WESTERN BULGARIA

Ivan Ts. Marinov, Tatiana StankovaForest Research Institute – SofiaBulgarian Academy of Sciences

Abstract

The paper presents results of investigation on water runoff and soil erosion at different land use types in Maleshevska Mountain, South-Western Bulgaria. The study is based on long-term (1995-2007) stationary data collection which included measurements of the water runoff and the eroded soil from experimental plots of different land use. The quantity of daily precipitations Q (mm) was recorded, the rainfall energy E(MJ/ha), the maximum of 30-minute rainfall intensity I30 (mm/min) and the precipitation indices RW = EI30 and R2 = QI30 were estimated based on it, and the influence of these principal characteristics of the precipitations on the water runoff and eroded soil was investigated.

The largest amounts of water runoff and eroded soil were recorded for the fallow lands and the tobacco field. The lowest values were established for the grass-covered plots and the oak stand. The precipitation indices RW and R2 were appropriate for distinguishing of different categories of runoffs: significantly higher amounts of water runoff and eroded soil were found for precipitations of high RW and R2 indices - about three times as much as those of the medium ones.

Positive and significant correlations between the water runoff and the eroded soil with general downward tendency with the increase of the plant cover were established.

Key words: erosion, eroded soil, precipitation indices, land use type

INTRODUCTION

The soil loss estimation precision is of major importance in designing anti-erosion activities in any region, field or watershed. Thus, evaluation of the main man-manageable (land management activities, grazing) and natural factors (climate, vegetation, relief, soils, main rock) is required.

The quantitative evaluation of the soil losses has been initially based on dependences of the solid runoff on the main erosion-driving factors. Later on, methodological approaches of more complex nature have been proposed, as it was the case with the Universal Soil Loss Equation (USLE) by Wischmeier, Smith (1978) and its revised version RUSLE (Renard et al., 1997). Numerous models for estimation of the erosion have been recently developed worldwide. Sufficient long-term data are required for model examination and adaptation as well as for investigation of the relationships between the precipitation and the water and the solid runoffs.

6

Precipitations, especially the intensive ones, are the main factor for soil erosion. Their influence on the soil erosion is usually characterized by indices formulated from the quantity (Q), maximum intensity (I), energy (E) and duration (t) of the rainfall (Wischmeier, Smith, 1978; Stanescu et al., 1969; Lal, 1976; Braunovic, 1996; Palecki et al., 2001). In the derivation of the Universal Soil Loss Equation Wischmeier, Smith (1978) defined the erosion index of a particular precipitation (Rw= EI30) on the basis of established relationship between the soil loss and the product of the maximum 30-minute precipitation intensity (I30) and its kinetic energy (Е). Dependencies of the quantities of the water and the solid runoffs on other index characteristics of the intensive rainfalls (R1 = QI15, R2 = QI30, R6 = QI10, Ro=Q/√t) have also been found (Stanescu et al., 1969; Mandev, 1976; Onchev, 1983; Rousseva, 2002) and subsequently applied for predicting the soil loss due to erosion. A number of studies, on the other hand, have shown the influence of the vegetation and the different ways of using land on the water runoff (Kostadinov, 1995; Olijnyk, 1999; Kulchytskyi-Zhyhailo, 2003; Dunjo et al., 2004, Oshurkevich, 2006; Konz et al., 2010).

The erosion processes in the mountainous regions of Bulgaria have been explored by numerous investigations (Kerenski et al., 1968; Kerenski, 1972; Biolchev et al., 1979; Angelov, 1986; Marinov, 1984; Kitin, 1988; Mandev, 1995, 1996, 1998; Marinov, 2007, 2009). Stationary observations of the hydrological and erosion processes on permanent plots and small watersheds have been carried out for more than 30 years (Raev, 1994, 2005; Marinov, 2009). The data recorded are applicable to study the relationships between the precipitation, the water runoff and the eroded soil, for model testing and for improvement of the research methodology.

The main objective of the present investigation is to evaluate and compare water runoff and soil erosion from runoff-inducing intensive precipitations at different land use types (oak forest, glade, pasture, fallow and tobacco lands) and to determine the influence of some characteristics of the precipitations on them on the basis of long-term stationary studies of permanent sample plots in Maleshevska Mountain, South-Western Bulgaria.

MATERIALS AND METHODS

Site descriptionThe study was carried out on the territory of ‘Igralishte’ ecological and erosion

investigation station situated in Sedelska river watershed basin in Maleshevska Mountain, South-Western Bulgaria. Four small watersheds of different land cover are set apart in the station area (Mandev, 1984; Marinov, 2009).

The watershed basin of Sedelska river is a part of Strumsko-Mestenska xerothermic zone of the South-Western Bulgaria. Previous studies identify the soils on the territory of the station as Luvisols (Mandev, 1984), which upper horizon has 1.31 to 1.34% humus content (Velizarova et al., 2010). The soils on the northern slope are characterized as relatively deep, of sandy to sandy-clay composition. The soils on the southern slope have clay-sandy to sandy composition and are shallow to very shallow and severely eroded.

7

The soil erodibility (K factor) vary from 0.028 to 0.044 (Mg/ha)[(MJ/ha)(mm/h)]-1, of average values 0.036 and 0.029 (Mg/ha)[(MJ/ha)(mm/h)]-1 for sunny and shady slopes, respectively. The annual precipitations for the period 1973-2007 in the region of station are 605 mm, as about 46% of them are of quantity more than 20 mm (Marinov, 2009). The amount of precipitations peaks in November-December, with second smaller peak in May-June, while the rainfall is the lowest in January and September. The annual precipitations during the period 1984-1994, i.e. 11 consecutive years, were less than the average annual quantity characteristic for the region, which led to formation of pronounced drought period in July, August and September. The mean annual number of intensive rainfalls (of at least 5-minute intensity exceeding 0.180 mm min-1) for the period 1976-2006 is 8.3, only 3 of which exceeding 20 mm of precipitation quantity. The mean annual temperature is 10.6 0C.

Data collectionSoil erosion monitoring can be carried out on-site (at plot level) and off site (at

sub-catchment and catchment level) (Hartanto et al., 2003). In this paper are presented the results from surface and solid runoff investigations at plot level. Data collection took place from 1995 to 2007 and included measurements of the water runoff (m3/ha) and the eroded soil (suspended material and sediments) (t/ha) from experimental plots of different land use established on southern (3 plots) and northern (5 plots) slopes (Table 1). The main types of land cover are oak forest, glade, pastures, tobacco field and fallow lands. The studies on the plots of fallow land and pasture have been performed in two trials in order to eliminate differences of the micro-environmental variation. Water and solid runoff were collected in the tanks after each precipitation.

Data of the water runoff and eroded soil caused by runoff-inducing intensive precipitations were analyzed. More than 100 runoff-inducing precipitations with quantity ≥ 9.5 mm were registered between 1995 and 2007. The quantity of daily precipitations Q (mm) of intensity I ≥10.8 mm h-1 (0.180 mm min-1) for 5, 10, 15 or 30 min were recorded. The rainfall energy E (MJ/ha) and the maximum of 30-min rainfall intensity I30 (mm min-1) were calculated and further used for estimation of the precipitation indices RW = EI30 and R2 = QI30.

Statistical analysesThe influence of the factors rainfall quantity (Q) and land use type (Plot type),

precipitation index R2 and Plot type, and precipitation index RW and Plot type on the water and solid runoffs were studied by Repeated Measures Analyses of Variance (RMANOVA) for data sets of 98, 68 and 66 measurements, respectively. Preliminary investigation on the influence of the plot replication on the amounts of runoffs was performed in order to define the factor Plot type. Prior to the RMANOVA, the data sets for Q, R2 and RW were classified into groups by two-step cluster analysis using the Log-Likelihood Distance as a distance measure and Schwarz’s Bayesian Information Criterion as a clustering criterion.

8

Bonferroni Multiple Range Test at p<0.05 for comparison of the means by factors followed the RMANOVA when significant influence of the studied factor was proven. Correlation between water and solid runoff was explored and analysed by experimental plots.

RESULTS AND DISCUSSION

Preliminary analysesStatistically significant influence of the plot replication was observed for both

the water and the solid runoff amounts (Table 2). Consecutively, each replication was considered as a different plot type and 7 plot types were designated and tested (Table 1).

The daily rainfall quantities for the period 1995-2001 varied between 8 and 77 mm and were clustered into 2 groups of moderate and high quantity precipitations. On the other hand, the precipitation index R2 had values from 0.16 to 48.84 mm (mm h-1)

and was classified into 3 categories of low, medium and high values. The rates of index RW varied between 35.19 and 1148.98 MJ mm ha-1h-1, which were clustered into 2 groups of moderate and high values (Table 3).

Influence of the land use type and the precipitation on the water runoffThe RMANOVA showed that all analysed precipitation characteristics affect

significantly the quantity of the water runoff (Tables 4, 5 and 6). Rainfall quantity of more than 40 mm caused 2 to 8 times higher amount of water runoff. The precipitations in the group of moderate quantity resulted in 1.83 to 51.63 m3/ha water runoff while the high quantity rainfall produced water runoff of 14.44 to 117.16 m3/ha (Fig. 1A).

Significantly higher (p=0.028) quantity of water runoff was caused by the precipitations of high value for the index R2: mean water runoff of 58.15 m3/ha vs. 21.45 m3/ha and 23.33 m3/ha for the medium and low values of R2, respectively (Fig. 1C). Water runoff of 1.29 to 56.16 m3/ha was recorded for the precipitations of moderate RW index, while significantly higher water runoff of 18.93 to 166.91 m3/ha was measured for the rainfalls of high RW index (Fig. 1E).

The highest value of the water runoff was recorded for the fallow lands (plots 1 and 2) and the tobacco field (plot 6). The lowest values of the dependent variable were achieved for the pasture, glade and oak stand plots (Fig. 1A, 1C, 1E). Low quantities of water runoff have been recorded for lands covered with herbaceous vegetation as well as oak forests, which is in agreement with the results of other investigations where insignificant amounts of water runoff and eroded soil from natural forest stands and grasslands were found (Litovchenko, 1986; Midriak, 1990; Rafailova, 2004; Dunjo et al., 2004). The results for the oak stands are comparable to those for the grasslands (Fig. 1A, 1C, 1E), because of the characteristics of the larger part of this forest type, which at this advanced age (around 100 years) and management system is characterized by low canopy closure (up to 0.4). Consequently, the rainfall water interception is weakly influenced by the tree crowns and the runoff decrease is mainly due to the impeding

9

influence of the herbaceous floor cover. Statistically significant Plot×R2 interaction was found (Table 5) due to change in the hierarchy of the plot means for the medium and the high R2 values. The fallow land of northern aspect and the tobacco field (Plots 2 and 6) showed highest water runoff values (Fig. 1C), which can be ascribed to the lack of permanent vegetation cover, the soil erodibility status and the direct impact of the rainfall on the ground surface. For the low and medium index R2 the tobacco field had highest value of water runoff, while for the precipitation with high R2 index an abrupt increase of the runoff for the fallow land, up to the maximum measured values, was recorded (Fig. 1C). The pasture on the slopes with southern aspect was affected by the increase in the rainfall quantity from moderate to high more severely than the glade and the oak stand, which caused statistically significant Plot×Q interaction (Fig. 1A). The steeper increase in the water runoff from the pasture on southern aspect with the increase in the RW precipitation index reflected in significant Plot× RW interaction (Fig. 1 E).

Influence of the land use type and precipitation on the eroded soilThe quantity of the eroded soil as a dependent variable proved strong influence

of the precipitation, characterized by Q, R2 and RW, and of the different land use types (Tables 4, 5 and 6).

The increased rainfall quantity caused doubling the mean amount of the eroded soil from the fallow lands and the tobacco field (Plots 1, 2 and 6) (Fig. 1B). The moderate quantity of precipitations resulted in 0.002 to 15. 274 t/ha solid runoff, while the large quantity of precipitations produced 0.006 to 33. 769 t/ha eroded soil.

A relationship between the rainfall quantity and the solid runoff is clearly distinguished which is in agreement with findings by other investigators (Roose, 1977; Arnoldus, 1978; Rousseva, 2002). Sirvent et al. (1997) established that eroded soil starts to increase after a certain threshold; above 15 mm of precipitation the sediments from uncovered soil clearly increased. Our investigation confirms that the largest quantity of precipitations do not necessarily produce the maximum soil erosion (Gonzalez-Hidalgo, 1994) because the recorded amounts of solid runoff vary within a broad range – up to 35.650 t/ha.

The precipitations of high index R2 resulted in three times increase in the eroded soil than those of low and medium precipitation index R2 (15.031 t/ha vs. 5. 888 and 5.085 t/ha, respectively). The average values of the water runoff and the eroded soil for the low index R2 were slightly higher than those for the medium index R2 (Fig. 1C, 1D), particularly for the plots of higher amounts of runoff. This might be due to the unequal number of cases in each of the sub-sets (Table 3), but also suggests that the precipitation index R2 is a rainfall characteristic less appropriate for the runoff investigation and classification, at least for these particular region as its values from 0.1 to 25 (more than 95% of the registered values) cause comparable runoff amounts.

Significantly higher amounts of eroded soil (0.005-54.432 t/ha) were recorded for the rainfalls of high RW index than those of moderate precipitation index RW (0.002-18.328 t/ha) (Fig. 1F). More intensive erosion was observed for the precipitations of higher energy (Onchev, 1983; Rousseva, 2002), which seems decisive for the strong

10

Table 1. Characteristics of the sample plots

Plot N/ Aspect Land use type Size,m × m

Plot area, m2

Slope, %

1) Southern Fallow land – Replicate 1 4×40 160 252) Northern Fallow land – Replicate 2 4×25 100 303) Southern Pasture – Replicate 1 4×40 160 254) Northern Pasture – Replicate 2 4×25 100 305) Northern Glade (grass cover >60%, without grazing) 4×60 240 24

6) Southern Tobacco field (tillage, planting and cultivation parallel to the contour lines) 4×25 100 25

7) Northern Oak forests, managed through branch-cutting (grass cover 80-90%) 5×30 150 32

Note: The size of all plots was unified to 100 m2 since 01.04.2001.

Table 2. Influence of the plot replication on the runoff amounts

Dependent variable

Factor

Water runoff Solid runoff

F-value Level of Significance F-value Level of

Significance

Land use type 66.059 P < 0.001 51.954 P < 0.001

Land use type × Replication 9.117 P < 0.002 9.547 P < 0.002

Replication 8.202 P < 0.005 9.392 P < 0.003

Table 3. Classification of the main studied precipitation characteristics

Factor group Numberof cases Mean value Standard

deviation Range

Q (mm)Moderate 79 23 8 8 – 39

High 12 56 10 43 – 77Rw =EI30 (MJ mm ha-1 h-1)

Moderate 59 140.02 82.54 35.19 – 350.69High 7 635.93 246.64 399.76 – 1148.98

R2=QI30 mm (mm h-1)Low 41 5.21 2.75 0.16 – 9.51

Medium 24 15.29 4.48 9.86 – 25.21High 3 40.30 9.59 29.93 – 48.84

11

Fig. 1. Mean values of the water runoff (A, C, E) and the eroded soil (B, D, F) by plots as influenced by the total daily precipitation Q (A, B), the precipitation index R2 (C, D) and Rw (E, F). Means with the

same letter are not significantly different at p<0.05 (Bonferroni multiple comparison test)

12

influence of the rainfalls of high RW on the formation of larger quantities of solid runoff. The fallow land of northern exposure appeared to be most susceptible to erosion (plot 2). The amount of solid runoff was more than 50.000 t/ha for the most intense and high quantity rainfalls (Fig. 1B, 1D, 1F). The small and similar quantities of eroded soil, regardless of the rainfall quantity for plots 4, 5 and 7 as compared to plots 1, 2 and 6, lead to manifestation of statistically significant Plot×Q interaction. The fallow land of southern exposure (plot 1) showed steeper increase in the quantity of the solid runoff with the increase of the precipitation index RW than the tobacco field (Fig. 1F) and the amount of the eroded soil was quite invariable for the oak stand regardless of the RW index. These events have probably resulted in the manifestation of statistically significant Plot× RW interaction.

The highest amount of solid runoff (eroded soil) was recorded for the open lands (the fallow land and the tobacco field), which is largely due to the soil operations involving the use of heavy machinery and ploughing techniques, which have a direct effect on soil losses, essentially increasing them (Nunes et al., 2011). The pronounced dry periods in July, August and September recorded during the last 11 years, followed by the season of the most intensive rainfalls, specific to the region, additionally increased the soil vulnerability and reinforced the soil loss processes. Almost the entire annual amount of the eroded soil (10-70 t/ha) is caused by 3 to 5 rainfalls. These are precipitations of quantity higher than 10 mm and of maximum 30-min intensity (I30) exceeding sometimes 50-70 mm/h. They belong to the second and the third group, in regard with the precipitation factor R2, since its value in these cases exceeds 9.86. Amount of eroded soil from the fallow land of more than 100 t/ha was recorded in two cases only. Similar results about the number of the precipitations, which cause most of the annual solid runoff, were reported also by González-Hidalgo et al. (2007). They found that soil erosion in Western Mediterranean areas depends on a few daily events – over 50% of soil eroded annually belongs to just three daily erosion events. The most severe soil erosion is triggered by intense rainfall and the percentage of precipitations that produce the greatest erosion is very low (López-Vicente et al., 2008).

The difference in the eroded soil between the fallow land (plot 1) (averages 11.583, 10.326 and 11.359 t/ha, respectively) and the tobacco field (plot 6) (averages 11.913, 12.047 and 12.717 t/ha, respectively) is minor (Fig. 1B, 1D, 1F), which is due to the same soil erodibility characteristics (Velizarova et al., 2010), but the eroded soil at the fallow land on northern slope (plot 2) is significantly greater which most probably reflects the influence of the slope gradient.

The oak stands, although exposed to branch-cutting and being much less dense, prevent the formation of large quantities of runoff and eroded soil, because of the dense grass cover formed under the oak canopy, which increases organic matter and reduces soil erodibility to ‘moderate’ (Velizarova et al., 2010). The insignificant amount of eroded soil from them showed that the processes taking place there are closer to the natural ones, rather than to those influenced by the human interference. Few studies of the erosion on steep slopes under forest have been conducted and the soil losses measured annually

13

Tabl

e 4.

Influ

ence

of t

he ra

infa

ll qu

antit

y Q

and

the

Plot

type

on

the

wat

er ru

noff

and

erod

ed so

il.

Dep

ende

nt v

aria

ble

Fact

or1

Wat

er r

unoff

Erod

ed so

il

Sum

of

squi

res

Deg

rees

of

free

dom

2M

ean

squi

reF-

valu

ean

d si

gnifi

canc

e3Su

m o

f squ

ires

Deg

rees

of

free

dom

Mea

n sq

uire

F-va

lue

and

sign

ifica

nce

Tests

of w

ithin

-sub

ject

effe

cts

Plot

type

3914

43.3

62.

7214

3786

.14

71.5

0***

3847

271.

271.

488

2585

754.

6346

.92*

**

Plot

type

× Q

6244

7.12

2.72

2293

8.26

11.4

1***

5651

83.3

01.

488

3798

60.2

36.

89**

Erro

r52

5587

.31

261.

3520

11.0

577

9014

7.85

141.

3555

113.

32

Tests

of b

etw

een-

subj

ect e

ffect

s

Q12

4759

.33

112

4759

.33

34.1

4***

4397

30.8

51

4397

30.8

59.

82**

Erro

r35

0864

.19

9636

54.8

442

5430

9.59

9544

782.

21

1 An

nota

tions

of t

he fa

ctor

s are

as d

efine

d in

the

text

2 Th

e de

gree

s of f

reed

om a

re re

calc

ulat

ed a

ccor

ding

to th

e G

reen

hous

e-G

eiss

er c

orre

ctio

n fo

r vio

latio

n of

the

sphe

ricity

ass

umpt

ion.

3 Si

gnifi

canc

e le

vel:

***

– P<

0.00

1; *

* –

P<0.

01

14

Tabl

e 5.

Influ

ence

of t

he p

reci

pita

tion

inde

x R 2 a

nd th

e Pl

ot ty

pe o

n th

e w

ater

runo

ff an

d er

oded

soil.

Dep

ende

nt v

aria

ble

Fact

or1

Wat

er r

unoff

Erod

ed so

il

Sum

of

squi

res

Deg

rees

of

free

dom

2M

ean

squi

reF-

valu

ean

d si

gnifi

canc

e3Su

m o

f squ

ires

Deg

rees

of

free

dom

Mea

n sq

uire

F-va

lue

and

sign

ifica

nce

Tests

of w

ithin

-sub

ject

effe

cts

Plot

type

1701

15.2

51.

877

9065

1.40

47.6

4***

1565

513.

301.

6694

3060

.77

32.9

7***

Plot

type

×R 2

2937

8.14

3.75

378

27.5

44.

11**

2089

05.4

13.

3262

922.

012.

20ns

Erro

r23

2103

.20

121.

978

1902

.83

3038

747.

4510

6.24

2860

2.08

Tests

of b

etw

een-

subj

ect e

ffect

s

R 225

688.

172

1284

4.08

3.77

*18

5677

.17

292

838.

584.

00*

Erro

r22

1353

.19

6534

05.4

314

8686

1.08

6423

232.

20

1 An

nota

tions

of t

he fa

ctor

s are

as d

efine

d in

the

text

2 The

degr

ees o

f fre

edom

are

reca

lcul

ated

acc

ordi

ng to

the

Gre

enho

use-

Gei

sser

cor

rect

ion

for v

iola

tion

of th

e sp

heric

ity a

ssum

ptio

n.3 Si

gnifi

canc

e le

vel:

***

– P<

0.00

1; *

* –

P<0.

01; *

– P

<0.0

5; n

s – P

>0.0

5

15

Tab

le 6

. Infl

uenc

e of

the

prec

ipita

tion

inde

x R W

and

the

Plot

type

on

the

wat

er ru

noff

and

erod

ed so

il.

Dep

ende

nt v

aria

ble

Fact

or1

Wat

er r

unoff

Erod

ed so

il

Sum

of

squi

res

Deg

rees

of

free

dom

2M

ean

squi

reF-

valu

e an

d si

gnifi

canc

e3Su

m o

f sq

uire

sD

egre

es o

f fr

eedo

mM

ean

squi

reF-

valu

e an

d si

gnifi

canc

e

Tests

of w

ithin

-sub

ject

effe

cts

Plot

type

2709

54.2

92.

4411

0949

.52

92.4

7***

2818

871.

981.

5018

8449

7.98

60.7

7***

Plot

type

× R

W59

642.

372.

4424

422.

1720

.35*

**72

4847

.42

1.50

4845

81.6

015

.63*

**

Erro

r18

7532

.71

156.

3011

99.8

529

2227

6.73

94.2

431

009.

95

Tests

of b

etw

een-

subj

ect e

ffect

s

R W10

1068

.49

110

1068

.49

43.7

4***

5399

28.8

01

5399

28.8

025

.06*

**

Erro

r14

7886

.54

6423

10.7

313

5738

2.57

6321

545.

76

1 An

nota

tions

of t

he fa

ctor

s are

as d

efine

d in

the

text

2 The

degr

ees o

f fre

edom

are

reca

lcul

ated

acc

ordi

ng to

the

Gre

enho

use-

Gei

sser

cor

rect

ion

for v

iola

tion

of th

e sp

heric

ity a

ssum

ptio

n.3 Si

gnifi

canc

e le

vel:

***

– P<

0.00

1.

16

usually do not exceed 1 t/ha (Morgan, Rickson, 1995). Mandev (1995) found for the area of the present study that on northern slope the annual amount of the eroded soil is 76.0 t/ha for fallow land, up to 1.17 t/ha for glade, 0.195 t/ha for the oak forests, managed through branch-cutting. On southern slope the annual amount of the solid runoff attains values of up to 123.5 t/ha (average 75.0) for the fallow land and the tobacco field and up to 37.0 t/ha (average 6.0) for the pasture.

Statistically significant interactions between the investigated rainfall characteristics and the land use type were distinguished for both types of runoff in almost all cases. This specificity in the rainfall influence implies the presence of critical thresholds of the studied precipitation parameters, which differ according to the land cover of the exposed soil.

Our results showed that the rainfall quantity Q and especially the precipitation index RW, can be used with high reliability in evaluation of the precipitation influence on the runoff of the studied region and their advantages are better distinguished in the study of the solid runoff. These characteristics can be used to model the hydrological processes of the studied land use types, because of the specificity of their influence distinguished through the highly significant Plot type×Q and Plot type×RW interactions.

Significant influence of the land use type and the plant cover on the runoff and soil erosion have been reported by numerous studies (Vacca et al., 2000; Dunjó et al., 2004; García-Ruiz, 2010; Mohamad, Adam, 2010; Cerdan et al., 2010; Nunes et al., 2011). Worldwide, erosion rates range from as low as 0.001–2.0 t/ha/yr on relatively flat land with grass and/or forest cover to rates ranging from 1 to 5 t/ha/yr on mountainous regions with regular vegetation cover (Pimentel, Kounang, 1998). Small amounts of eroded soil have been recorded, based on monthly erosion during the growing season, in the studies by Konz et al. (2010) from pasture without and with dwarf shrubs – up to 68 kg/ha and up to 11 kg/ha respectively. De Vente, Poesen (2005) reported the mean sediment production for various plots in Southeast Spain, which for the plots from 15 to 80 m2 were up to 810 t/km2/year, but for 25 of the 40 observations the eroded soil quantity was up to 50 t/km2/year, and the quantity for 60% of them was less than 10 t/km2/year. The same authors reported for the Mediterranean comparable average soil loss rates, between 0.8 and 142.8 t/km2/year, under diverse land use conditions and with plot sizes between 20 and 166 m2.

Table 7. Correlations between the water runoff and the eroded soil

Plot 1 Plot 2 Plot 3 Plot 4 Plot 5 Plot 6 Plot 7

Pearson correlation coefficient 0.854 0.346 0.378 0.607 0.810 0.464 0.248

Significance of the correlation <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.011

Number of cases 104 104 104 103 104 103 104

17

Correlations between the water runoff and the amount of the eroded soilPositive and significant correlations were established between the water runoff

and the eroded soil (Table 7). The correlations showed general tendency of decreasing with the increase of the plant cover and had lowest value (under 0.30) for the oak forests, managed through branch-cutting.

CONCLUSIONS

The largest amount of water runoff and eroded soil was determined from the fallow lands and the tobacco field and the lowest values – for the grass covered plots and in the oak stand. Significant influence of the precipitations on the amounts of the runoff and the eroded soil has been established. The highest amounts of eroded soil (up to 54. 43 t/ha) were recorded for the rainfalls of high RW index and the highest precipitation quantities caused up to 33.77 t/ha of eroded soil.

The results showed that the rainfall quantity Q and especially the precipitation index RW, can be used with high reliability in evaluation of the precipitation influence on the runoff of the studied region and their advantages are better distinguished in the study of the solid runoff. These characteristics can be used to model the hydrological processes and to control the erosion, because of the specificity of their influence according to the land use type, suggesting the existence of critical threshold values of these characteristics according to the land cover. The precipitation index R2, on the other hand, showed less appropriate for the runoff investigation and classification, because more than 95% of its registered values caused comparable runoffs amounts.

Positive and significant correlations between the water runoff and the eroded soil with general downward tendency with the increase of the plant cover were established.

REFERENCES

Angelov, S. 1986. On certain elements of the water balance of coniferous plantations and oak forests. Forest Science, Sofia, 23, 63-67 (In Bulgarian, English summary).

Arnoldus, H. J. 1978. An approximation of the rainfall factor in the Universal Soil Loss Equation. – In: De Boodt, M. and Gabriels, D. (Eds.) Assessment of erosion, Wiley, Chichester, 127-132.

Biolchev, A., P. Pimpirev, B. Kitin. 1979. Hydrological studies at the experimental area of ‘G. S. Avramov’ Learning and experimental Forestry Enterprise. Scientific works of the Higher Forestry Engineering Institute 18, Series Forestry, 163-171 (In Bulgarian).

Braunović, S. 1996. Erositivity of Rain in the Surroundings of Belgrade. – In: Proceedings of the First European Conference and Trade Exposition on Erosion Control. Barcelona, May 1996, IECA, Sitges – Barcelona, 29-31.

Cerdan, O., G. Govers, Y. Le Bissonnais, K. Van Oost, J. Poesen, N. Saby, A. Gobin, A. Vacca, J. Quinton, K. Auerswald, A. Klik, FJPM. Kwaad, D. Raclot, I. Ionita, J. Rejman, S. Rousseva, T. Muxart, M.J. Roxo, T. Dostal. 2010. Rates and spatial variations of soil erosion in Europe: A study based on erosion plot data. Geomorphology, 122, 167-177.

De Vente, J., J. Poesen. 2005. Predicting soil erosion and sediment yield at the basin scale: Scale issues and semi-quantitative models. Earth-Science Reviews, 71, 95-125.

Dunjo, G., G. Pardini, M. Gispert. 2004. The role of land use–land cover on runoff generation and sediment

18

yield at a microplot scale, in a small Mediterranean catchment. J. of Arid Environments, 57, 99-116. doi:10.1016/S0140-1963(03)00097-1.

García-Ruiz, J. M. 2010. The effects of land uses on soil erosion in Spain: A review. Catena 81, 1-11. doi:10.1016/j.catena.2010.01.001.

González-Hidalgo, J. C. 1994. Efectos de la superficie protegida en la variación espacial de la producción de sedimentos. Cuaternario y Geomorfología 8, 65-76.

González-Hidalgo, J. C., J. L. Penna-Monne, M. De Luis. 2007. A review of daily soil erosion in Western Mediterranean areas. Catena, 71, 193-199.

Hartanto, H., R. Prabhu, A. S. E. Widayat, Ch. Asdak. 2003. Factors affecting runoff and soil erosion: plot-level soil loss monitoring for assessing sustainability of forest management. Forest Ecology and Management, 180, 361-374.

Kerenski, S. 1972. Study on the water-regulating and anti-erosion function of a riverside Austrian pine plantation. PhD Thesis. Forest Research Institute of BAS, Sofia, 160 (In Bulgarian). Kerenski, S., S. Angelov, A. Mandev. 1968. On the Erosion by Drakovska River and its Control. Forest Science, Sofia, 4, 47-64 (In Bulgarian).

Kitin, B. 1988. Studies on the hydrological and erosion controlling role of coniferous forests. PhD Thesis. University of Forestry, Sofia, 417 (In Bulgarian).Konz, N., D. Baenninger, M. Konz, M. Nearing, C. Alewell. 2010. Process identification of soil erosion in steep mountain regions. Hydrology and Earth System Sciences, 14, 675–686. doi:10.5194/hess-14-667-2010.

Kostadinov, S. 1995. Hydrological regime in a torrential watershed of a hilly-mountainous region of South-East Serbia. Lesnictvi-Forestry, 41, 285-292.

Kulchytskyi, Z. 2003. The runoff regulative function of forests on mountain river watershed. In: Hamor, F., Commarmot, B. (Eds.) Natural Forests in the Temperate Zone of Europe – Values and Utilization. Carpathian Biosphere Reserve – Rakhiv and the Swiss Federal Research Institute – Birmensdorf, Ukraine, 210-211.

Lal, R. 1976. Soil erosion on Alfisols in Western Nigeria. III. Effects of rainfall characteristics. Geoderma, 16, 389-401.

Litovchenko, A. F. 1986. Experimental study of the elements of water balance in mountain watersheds. Highest school. Import. Publish., Kiev, 185 (In Russian).López-Vicente, M., A. Navas, J. Machin. 2008. Identifying erosive periods by using RUSLE factors in mountain fields of the Central Spanish Pyrenees. Hydrology and Earth System Sciences, 12, 523-535. doi:10.5194/hess-12-523-2008.

Mandev, A. 1976. Influences of Certain Factors upon the Soil Erosion in the Region of Sandanski. Dissertation, Forest Research Institute of BAS, Sofia (In Bulgarian).

Mandev, A. 1984. A study on the water and solid flow in little catchments in Malashevska mountain. − Forest Science, Sofia, 3, 45-64 (In Bulgarian).

Mandev, A. 1995. Regularities in the variation of the intensity of sheetwater erosion in upland catchment areas managed in a number of ways. Scientific Conference with Participation of Foreign Specialists ‘90 Years of Soil Erosion Control in Bulgaria’, Sofia, 37-41 (In Bulgarian).

Mandev, A. 1996. Evaluation of the soil protective effectiveness of some forest ecosystems in Southwestern Bulgaria. – In: Proceedings of the Second Balkan Scientific Conference: Study, Conservation and Utilization of Forest Resources II, 43-47.

Mandev, A. 1998. Assessment of the hydrological functions of the some forest ecosystems in South-west Bulgaria. – In: Proceedings of the Scientific conference with international participants ‘70 years Forest Research Institute’, 6-7 October 1998, Sofia, 94−98 (In Bulgarian).

Marinov, I. Ts. 1984. A Study on Water-Regulating and Soil-Protecting Role of the Forest Stands in the Watershed of the River Melnishka. Forest Science, Sofia, 2, 22-38 (In Bulgarian).

Marinov, I. Ts. 2007. Investigation on the maximum quantities of water runoff and eroded soil from differently managed hill-side lands. Forest Science, Sofia, 4, 47-58 (In Bulgarian).

Marinov, I. Ts. 2009. Investigation and Evaluation of Erosion in Some Regions of Southwestern Bulgaria. DSc Thesis, Forest Research Institute – BAS, Sofia (In Bulgarian).

Mohammad, A. G., M. A. Adam. 2010. The impact of vegetative cover type on runoff and soil erosion under different land uses. Catena, 81, 97-103. doi:10.1016/j.catena.2010.01.008.

19

Midriak, R. 1990. Surface runoff and soil loss in alpine ecosystems. Scientific Works, Research Institute of Forestry in Zvolen, Slovakia, 39.

Morgan, R. P. C., R. J. Rickson (Eds.). 1995. Slope Stabilization and Erosion Control – A Bioengineering Approach. E & F Spon, London, 274.

Nunes, A. N., A. C. De Almeida, C. O. A. Coelho. 2011. Impacts of land use and cover type on runoff and soil erosion in a marginal area of Portugal. Applied Geography, 31, 687-699. doi:10.1016/j.apgeog.2010.12.006.

Olijnyk, V. S. 1999. Flood control role of the mountains forests in Carpathian Mountains, Ecological and social-economic aspects of the catastrophic phenomenon in Carpathian region. Intern. scient.-pract. conference, Uzhgorod, Carpaty, 257-261.

Onchev, N. 1983. Prognostication of the Surface Water Erosion in the People’s Republic of Bulgaria and Optimisation of the Anti-Erosion Operations, Monograph. Agricultural Academy, Sofia. 204 (In Bulgarian).Oshurkevich, O. 2006. The water runoff of mountain catchments of different area and percentage of forest. – In: Boczon A (Ed.) Assessing of soil and water conditions in forests, Forest Research Institute, Warsaw, 31-38.

Palecki, M. A., J. R. Angel, S. E. Hollinger. 2001. Storm precipitation structure in the Eastern United States. Part I: spatial changes, 12th Symposium on Global Change and Climate Variations, Joint with the Symposium on Precipitation Extremes: Predictions, Impacts, and Responses, 14-18 January, 2001, Albuquerque, New Mexico. Boston, MA: American Meteorological Society, paper J2.2, 59-60.

Pimentel, D., N. Kounang. 1998. Ecology of Soil Erosion in Ecosystems. Ecosystems, 1, 416-426.Raev, I. 1994. Certain results of the years-long stationary investigations carried out by the Forest Research

Institute in the Bulgarian Academy of Sciences in forests in Bulgaria. – In: Proceedings of National conference of forest protection and monitoring of forest ecosystem. BAS, 149-180 (In Bulgarian).

Raev, I. 2005. Research network, achievements and problems of the forest hydrological studies in Bulgaria. – In: Proceedings of the International Conference ‘Forest Impact on Hydrological Processes and Soil Erosion’, 5–8 October 2005, Yundola, Bulgaria, 17-29.

Rafailova, E. 2004. Impact of forestry activities on hydrological processes in mixed coniferous forests. PhD Thesis, University of forestry, Sofia (In Bulgarian).

Renard, K. G., G. R. Foster, G. A. Weesies, D. K. McCool, D. C. Yoder. 1997. Predicting soil erosion by water – A guide to conservation planning with the revised universal soil loss equation (RUSLE). U.S. Department of Agriculture, Agruculture Handbook 703, 384.

Roose , E. J. 1977. Application of the Universal Soil Loss Equation of Wischmeier and Smith in West Africa. – In: Greenland, D.J., Lal, R. (Eds.), Soil and Water Conservation and Management in the Humid Tropics, Wiley, Chichester, UK, 177-187.

Rousseva, S. S. 2002. Information Bases for Developing a Geographic Database for Soil Erosion Risk Assessments. Monograph. Nikola Poushkarov Institute of Soil Science, Sofia, 198.

Raev, I. 2005. Research network, achievements and problems of the forest hydrological studies in Bulgaria. - In: Proceedings of the International Conference ‘Forest Impact on Hydrological Processes and Soil Erosion’, 5-8 October 2005, Yundola, Bulgaria, 17-29.

Sirvent, J., G. Desir, M. Gutierrez, C. Sancho, G. Benito. 1997. Erosion rates in badland areas recorded by collectors, erosion pins and profilometer techniques (Ebro Basin, NE-Spain). Geomorphology, 18, 61-75.

Stanescu, P., I. Taloiescu, L. Dragan. 1969. Contribuţii in studierea unor indicatori de evaluare a erozivitatii pluviale. GAnuarul I. C. P. A. Seria Pedologie, II (XXXVI), Bucureşti, 361-369.

Vacca, A., S. Loddo, G. Ollesch, R. Puddu, G. Serra, D. Tomasi, A. Aru. 2000. Measurement of runoff and soil erosion in three areas under different land use in Sardinia Italy. Catena, 40, 69-92.

Velizarova, E., I. Ts. Marinov, T. Lubenov. 2010. Assessment of Soil Erodibility of Different Land-Use Types in a Small Mountain Watershed. – In: M. Zlatic (Ed.), Global Change – Challengers for Soil Management, Advances in GeoEcology, 41, ISBN 987-3-923381-57-9, 108-118.

Wischmeier, W. H., D. D. Smith. 1978. Predicting Rainfall Erosion Losses. A guide to Conservation Planning. USDA Agr. Handbook 537, US Department of Agriculture, 58.

E-mail: marinovbg02@yahoo.com