Sa

NC

a

ARRAA

KSCACISSS

1

dilItlo

N

I

0h

Agriculture, Ecosystems and Environment 158 (2012) 31– 40

Contents lists available at SciVerse ScienceDirect

Agriculture, Ecosystems and Environment

jo ur n al homepage: www.elsev ier .com/ lo cate /agee

oil carbon sequestration and erosion control potential of hedgerowsnd grass filter strips in sloping agricultural lands of eastern India

arendra K. Lenka ∗, Anchal Dass1, S. Sudhishri1, U.S. Patnaikentral Soil and Water Conservation Research and Training Institute, Research Centre, Sunbeda, Koraput, Orissa, India

r t i c l e i n f o

rticle history:eceived 28 February 2012eceived in revised form 16 May 2012ccepted 19 May 2012vailable online 17 June 2012

eywords:oil restorationarbon sequestrationgroforestryontour hedgerows

n situ soil moisture conservationoil qualityustainable agricultureoil conservation

a b s t r a c t

Contour hedgerows and grass filter strips are important towards enhancing and sustaining productivityof sloping agricultural lands in medium to high rainfall regions. However, impact of such measures onerosion control, soil carbon sequestration and agronomic productivity have not been widely assessed forthe small land holders in eastern India. Therefore, an on-farm study was conducted between 2001 and2006 to evaluate the impact of the techniques on soil organic carbon (SOC) concentration and pool; lossesof water, soil and nutrients; soil moisture storage and agronomic yield on arable lands of 2–5% slope. Thestudy was taken in 5.95 ha area with six treatments and nine replications. Treatments consisted of twohedgerow species (Gliricidia sepium and Indigofera teysmanni) and a control, with or without grass filterstrip (GFS) of a local species (Saccharum spp.). Using finger millet (Eleusine coracana) as the test crop,the hedgerow species were planted at 0.5 m × 0.5 m spacing in staggered double rows and the GFS ina single row at 0.3 m spacing. In general, Gliricidia + GFS was most conservation effective followed byIndigofera + GFS. It reduced runoff by 33% (10.7% runoff compared to 16.1% in control), soil loss by 35%(6.3 Mg ha−1 compared to 9.71 Mg ha−1 in control), and SOC loss through runoff by 50 kg ha−1 yr−1. Inaddition, it resulted SOC build up at 0.352–1.354 Mg ha−1 yr−1 at three graded distance from hedgerows,out of which 0.352 Mg ha−1 yr−1 was sequestered due to soil reclamation and about 1.0 Mg ha−1 yr−1 wasretained due to barrier effect. With higher soil moisture storage by 28–37 mm and 22–43 mm at 12 and17 days of dry spell, respectively, the grain yield of finger millet increased by 49% from 952 kg ha−1 incontrol to 1413 kg ha−1 in Gliricidia + GFS treatment. Addition of GFS significantly reduced the losses of

−1

water runoff, soil and nutrients in all the treatments, and increased SOC stock by 0.38–1.0 Mg ha inthe 0.6 m soil profile. The GFS also improved soil moisture storage by 9–12 mm and 6–15 mm at 12 and17 days of dry spell, respectively. As compared to the pre-treated initial, the SOC stock decreased by60–112 kg ha−1 yr−1 in the control indicating on-going erosion process in unprotected lands. The studyshowed the C sink potential of erosion control measures in the sloping agricultural lands of eastern India.

. Introduction

Decline in soil quality, depleting soil organic carbon (SOC) andegradation of land resources due to erosion are the major imped-

ments for future global food security. The productivity of someands has declined by 50% due to soil erosion and desertification.n South Asia, annual loss in productivity is estimated at 36 million

ons of cereal by water erosion (Eswaran et al., 2001). Eroded landseft unprotected lead to further erosion on-site and have greaterff-site impacts. On the other hand, rehabilitation of eroded lands

∗ Corresponding author. Present address: Indian Institute of Soil Science,abibagh, Bhopal, MP, India. Tel.: +91 755 2730970; fax: +91 755 2733310.

E-mail address: nklenka@rediffmail.com (N.K. Lenka).1 Present address: Indian Agricultural Research Institute, New Delhi 110012,

ndia.

167-8809/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.agee.2012.05.017

© 2012 Elsevier B.V. All rights reserved.

with conservation measures not only reverses the process of soildegradation but also improves the soil quality and converts theselands to potential carbon sinks (Lal, 2008; Lenka et al., 2012).

In India, 146.82 million ha (about 45% of the land area) areais under various forms of land degradation (SoER - India, 2009).Degradation is particularly severe in regions with sloping and hillyterrains and those affected by unsustainable land managementpractices such as shifting cultivation. The sloping and hilly regionsof eastern India, called eastern ghats region with a geographical areaof 19.8 million ha (Sikka et al., 2000) is such an erosion prone zone,having characteristic link of poor lands with people’s poverty. Forinstance, the share of good quality soil in Orissa is one of the low-est, merely 10.4% of the land area of the state (Kumar, 2011). It also

happens to be the most backward state of India with 46.4% of thepopulation below poverty line (Planning Commission, GoI, 2012).

Shifting cultivation is prevalent in the hill slopes of the region.However, reduction in restoration or fallow cycle from 15 to 20

3 stems

yr(twthsssoculdttss

aOgtwlafhpsff2aamiAaTtg

1tietofscts2l2tftet

ap

2 N.K. Lenka et al. / Agriculture, Ecosy

ears to the current level of 2–3 years due to population pressure,esulted in reduced farm output and increased land denudationLenka et al., 2012). This shifted focus of the people to settled cul-ivation on the sloping and undulated uplands and medium lands,ith average slope varying from 2 to 5% and characterized by coarse

extured Alfisols. These lands are located downside the denudedillocks and are the major alternatives for the predominantly sub-istence agriculture practiced in this rainfed region of India. Lowoil fertility and erosion due to overland flow from denuded hilllopes do not permit more than one crop per year and a crop yieldf more than 1.0 Mg ha−1 in these lands. Finger millet (Eleusine cora-ana L.) is the most common crop and in lands with better fertility,pland rice (Oryza sativa L.) is grown. Left unprotected, these arable

ands yield to high runoff, get eroded and further damages theownstream cultivated lands due to erosion and siltation. Puttinghese lands to conservation treatments can restore productivityhrough reduced soil and nutrient loss, increased soil moisture andoil organic carbon (SOC) storage and consequent improvement inoil quality.

Mechanical measures for controlling soil erosion are not afford-ble by individual farmers because of extreme poverty condition.n the contrary, vegetative measures involving hedgerows andrasses are cost-effective, durable and find people’s acceptance inhis region as they offer multiple benefits such as for fodder and fuelood. They are effective in low to medium slope ranges of arable

ands (Chunale, 2004; Dass et al., 2011). The species generally usedre vegetative barriers of grass species or shrubs and their per-ormance for soil and moisture conservation depends upon theiredge forming ability (Sharma et al., 2002). Hedgerow intercrop-ing though initially developed to restore the fertility of degradedoils in the tropics has been adopted in other regions not onlyor soil amelioration, but also to provide additional products (e.g.odder) and services (e.g. erosion control) (Albrecht and Kandji,003). Contour hedgerows are reported to promote the SOC stor-ge because of a local effect under the hedge and also due to theirnti-erosive effect (Walter et al., 2003). They are also effective inaintaining soil fertility and reducing the soil and nutrient losses

n sloping lands (Lin et al., 2009; Tao et al., 2012; Xu et al., 2012).s the cultivated lands are scarce and fragmented, systems such aslley cropping are not popular in arable lands of the study region.he most acceptable measures are modification to field bundshrough strengthening with vegetative measures with shrubs orrass species.

Management practices such as conservation tillage (Lal et al.,999; Schlesinger, 2000) and erosion control measures can improvehe SOC stock and net C sink potential of sloping arable lands. Keep-ng in view the finite C sink capacity of soil (Chung et al., 2010),roded lands, if put under erosion control measures, can be poten-ial C sinks. Certain soil management practices such as applicationf manures, fertilizers, irrigation of semi-arid and marginal landsor crop production, though increase the SOC status, are not net Cinks for CO2 emission and do not contribute to the Kyoto Proto-ol because of the associated carbon costs (Schlesinger, 2000). Evenhe advantages of no-till system over conventional tillage for SOCequestration is questioned in recent studies (Blanco-Canqui et al.,011; Ogle et al., 2012). SOC build up may be higher where the

and cover is fully changed to pasture or agroforestry (Saha et al.,010; Lenka et al., 2012). But, subsistence farming as prevalent inhe region, (Srivastava et al., 2004) may not permit pasture or agro-orestry in agricultural lands used for growing food crops, even ifhey are eroded. On the other hand, keeping the land use unaltered,roded lands can be treated with conservation measures to offset

he on-site and off-site impacts on soil and environment.

Much of the studies on SOC storage and C sequestration inrable lands have focussed on tillage and residue managementractices, but not on erosion control measures. Even, in majority of

and Environment 158 (2012) 31– 40

sequestration studies, the sampling depth is restricted to 30 cm orless (Baker et al., 2007), which gives an unclear picture about theeffects of conservation tillage on C sequestration. On the contrary,erosion control measures apart from soil amelioration effect mini-mize the loss of C by reducing the runoff and soil loss, which shouldbe counted while computing the net C sink benefits. However,impact of such measures on erosion control, soil carbon seques-tration and agronomic productivity have not been widely assessedfor the small land holders in eastern India. Thus, this study attemptsto compare the net soil C sink potential of selected soil conservationtreatments and their efficacy for retaining nutrients and moisturein the sloping arable lands of a sub-humid sub-tropical region ofeastern India.

2. Materials and methods

2.1. Experimental site

The experiment was conducted in farmers’ fields in a participa-tory research mode for 5 consecutive years during 2001–2005 in amicro-watershed located at Kokriguda village in Koraput district ofOrissa, in the eastern ghats of India (Fig. 1). The selected watershedis a completely tribal village with illiterate (less than 10% literacy)and poor populace. The experiment site is located at 18◦ 45′ N lati-tude and 82◦42′E longitude and at 910 m above mean sea level. Thestudy area comes under sub-tropical and sub-humid type of cli-mate, with annual mean maximum and minimum temperatures of30.6 ◦C and 17.0 ◦C, respectively. Mean annual rainfall is 1373 mm,80% of which is received during June–September. Soil type is pre-dominantly red lateritic and acidic with pH around 6.0 and comeunder udic paleustalfs as per USDA soil classification.

2.2. Experimental treatments

The experimental treatments consisted of selected soil con-servation systems on miniature field bunds of 0.15 m × 0.60 m(height × width) in 54 plots in farmers’ fields with two hedgerowspecies – Gliricidia sepium and Indigofera teysmanni – integratedwith or without filter strips of a local grass species – Sambuta(Saccharum spp.); a control and a sole grass filter species (GFS),taken in nine replicates in a randomized block design. Thus, thetreatments were: (1) Gliricidia, (2) Gliricidia + GFS, (3) Indigofera,(4) Indigofera + GFS, (5) Control, (6) Sole GFS. The treatments cov-ered 5.95 ha area, with average land slope of 2–5%. The minimumrunning length of the conservation system for each treatmentwas 10 m. The hedgerow species were planted at a spacing of0.5 m × 0.5 m in staggered double rows and the grass filter specieswas planted in a single row at 0.3 m. A total of 1800 m bund lengthwas planted with the selected treatment species during July 2001.A third hedgerow species, perennial arhar (Cajanus cajan) thoughwas taken initially but was a failure due to its poor survival. Thus,from the year 2002, the treatments were continued with two hedgerows species and a grass species, as mentioned above, with a con-trol and the monitoring of data such as runoff and soil loss startedfrom 2002. A schematic presentation of the treatment lay out isgiven in Fig. 2.

2.3. Test crop

To compare the efficacy of conservation treatments on crop per-formance, a medium duration (105 days) finger millet crop (cv.Bhairabi) was taken at a row–row spacing of 0.2 m across the slope

in the first week of July during the experiment period. As the experi-ment was conducted on a participatory on-farm research mode, thecrop was sown direct seeded under rainfed conditions with nutrientand weed management practices followed by the local farmers. The

N.K. Lenka et al. / Agriculture, Ecosystems and Environment 158 (2012) 31– 40 33

Fig. 1. Location of the study area.

0.3 m

0.5 m

Hedge row (1st row)

Slope

Sampling point at2.0 m from bun d

Sampling point at1.0 m from bun d

Crop are a

0.5 m

Hedge row (2nd row)

Grass filter strip

0.5 mField bun d

Fig. 2. A schematic diagram showing components of soil conservation treatments on field bunds.

3 stems

cphca

2

pefmars

2

1afptmtstgw

2

miwaatssd6wrrapvs

2

wwGb1ro13s

4 N.K. Lenka et al. / Agriculture, Ecosy

rop was supplied with 25 kg ha−1 each of urea and di-ammoniumhosphate and 2.0 Mg ha−1 of FYM. Weeds were controlled by oneand weeding at 30 days after sowing and without use of any herbi-ide. Yield of test crop was monitored for 2 consecutive years (2004nd 2005).

.4. Growth performance of hedgerow and grass filter species

To compare the growth performance of treatments, bio-physicalarameters such as survival, plant height, collar girth, canopy diam-ter, height at first branch and number of branches were recordedor the hedge species every year in the last week of October, up to 30

onths after planting. After 30 months, hedge species were prunednd maintained at 0.5 m height. For the grass filter species, biomet-ic observations such as survival, clump diameter and number oflips were recorded up to 30 months after planting.

.5. Soil sampling

Soil sampling was done by a screw auger from 0 to 15 cm,5–30 cm and 30–60 cm soil depths at two graded distance (1.0 mnd 2.0 m) from the treated bunds in the up-stream direction andor the whole plot. The samples of 1 m and 2 m bund distance wererepared from composite samples of two auger points to minimizehe land gradient effect. The whole plot samples were prepared by

ixing soils of four auger points including one from centre andhree from the plot leaving 2 m gap from bunds in each side. Initialoil samples were collected in 2001 and after 5 years during Mayo June of 2006. The samples were air dried under shade and thenround to pass through a 2 mm sieve. For soil bulk density, samplesere collected by using soil cores from each treatment.

.6. Monitoring of runoff, soil and nutrient loss

Monitoring of runoff, soil and nutrient loss was carried out byulti-slot divisors installed in only one replication of the exper-

ment. The plots for runoff studies were 8 m wide and 60 m longith two cross bunds having the selected conservation treatments

t a horizontal distance of 20 m. Runoff collected was measuredfter rainy days with rainfall of more than 12.5 mm, by measuringhe depth of water collected in the runoff collection tanks. Runoffamples were collected for soil and nutrient loss estimation. Theamples were allowed to settle down and the supernatant wasecanted and the settled sediments were dried in hot air oven at5 ◦C to estimate the soil loss. The runoff and soil loss monitoringas done for 3 consecutive years from 2002 to 2004. The required

ainfall data was collected from the nearby rain gauge station. Totalainfall of 585, 717, 771, and 1096 mm was received in 37, 49, 48nd 66 rainy days during the experiment period of 2002–2005. Thean evaporation data was collected from the meteorological obser-atory of CSWCRTI Research Centre, about 16 km from the studyite.

.7. Monitoring soil moisture

In the 5th year of the study (2005–2006), impact on soil moistureas studied at certain days’ interval from last week of September,hen the dry period started due to cessation of monsoon rainfall.ravimetric samples were collected at two graded distance fromunds (1 m and 2 m) and from the centre of each plot, from 0 to5 cm, 15–30 cm and 30–60 cm soil depths. In case of interveningainfall of more than 2.5 mm, samples were collected after 4 days

f rainy day and then at 7 days interval. Two such periods with2 and 17 days of dry spell with corresponding evaporativity of4 and 47 mm, were chosen for comparison purpose. Volumetricoil moisture storage up to 0.6 m soil depth was computed from

and Environment 158 (2012) 31– 40

the gravimetric soil moisture data and the depth wise bulk densityvalues.

2.8. Soil analysis

The soil bulk density was determined by core method (Black,1965). Soil organic carbon was determined by Walkley and Black’swet-digestion method (Jackson, 1973). Available N, P and K in thesediments were determined by alkaline KMnO4 method, Bray – Imethod and NH4OAC method, respectively (Jackson, 1973).

2.9. Soil organic carbon stock, SOC sequestration rate and buildup rate

Soil organic carbon stock was computed on constant depth basisup to 0.6 m profile for incremental soil layers (0–15 cm, 15–30 cmand 30–60 cm) by taking the organic carbon and bulk density datafor each layer. Effect of a particular treatment was assessed in termsof carbon build up rate by subtracting the initial carbon stock valuesfrom that of selected year and by dividing the time gap.

SOC0.6 m =n∑

i=1

SOCi × BDi × Di × 100

where SOC0.6 m is the soil carbon stock over 0–0.6 m depthexpressed in Mg C ha−1, SOCi is the soil organic carbon concentra-tion (Mg C 100 Mg−1 soil) for depth interval i (Di in m), BDi (Mg m−3)is the bulk density for depth interval i.

SOC build up rate : Ct = (Cx − Cx−n)n

where Ct = rate of soil carbon build up (Mg ha−1 yr−1); Cx = SOC stockin the Xth year; Cx−n = SOC stock in n years before Xth year.

Soil C sequestration rate is generally computed by comparingthe SOC stock in a particular year to that of the initial (Srinivasaraoet al., 2011). However, comparison with untreated (control) andpre-treated (initial) mean differently particularly for eroded lands.This is because of on-going process of erosion and continuing soiland nutrient loss in untreated (control) plots. It may be possi-ble that the SOC status of the pre-treated (initial) is higher thanuntreated (control). Thus, in this study, SOC sequestration rate wascomputed by comparing the SOC stock in the particular year to thatof the untreated (control), as per the following formula. Total SOCsequestered was the difference between the SOC stock in a givenyear and that in control.

SOC sequestration rate : Csr = (Cct − Cc)n

where Csr = SOC sequestration rate (Mg ha−1 yr−1); Cct = SOC stockin the conservation treatment; Cc = SOC stock in control; n = numberof intervening years between the initial and the year of comparison.

2.10. Statistical analysis

The data were analyzed using analysis of variance (ANOVA) andthe results compared at 95% significance as relevant for a random-ized block design on the basis of least significant difference (LSD0.05) values (Cochran and Cox, 1963). The effect of treatments on

runoff and soil and nutrient loss was tested using paired t-test andthe mean values were compared at 95% significance level. The coef-ficient of variation (CV) was calculated by the ratio of ‘square rootof error mean square’ to ‘mean’, multiplied by 100.

N.K. Lenka et al. / Agriculture, Ecosystems and Environment 158 (2012) 31– 40 35

Table 1Growth performance of hedgerow species under different treatments after 30 months of planting.

Treatment Survival (%) Height (cm) Collar girth (cm) Canopy diameter (cm) Height of first branch (cm) No. of branches

0.25 m height Top

Indigofera 39.08 73.17 6.81 17.72 42.95 68.24 6.06Indigofera + GFS 41.68 71.40 6.78 17.35 40.92 77.06 7.07Gliricidia 56.06 116.88 9.12 23.88 73.62 89.38 5.20Gliricidia + GFS 51.49 101.42 8.86 22.26 60.19 87.65 6.63

2018

3

3

ttwrwGslowhcPbbGvpSr

fissf

3

I(rlmsoi

TR

Mean 47.08 90.72 7.90

LSD (0.05) 8.80 11.31 0.68

CV, % 19.48 12.46 9.04

. Results and discussion

.1. Growth performance of hedgerow and grass filter species

The performance of Gliricidia with or without GFS was betterhan Indigofera with or without GFS in most of the growth parame-ers (Table 1). Highest survival of 56% was observed under Gliricidiahich was at par with Gliricidia + GFS, but was significantly supe-

ior to both the treatments of Indigofera. Maximum plant heightas observed under Gliricidia which was significantly higher thanliricidia + GFS. Both treatments of Indigofera were at par, but withignificantly lower height than other treatments. In terms of col-ar girth and canopy diameter at 0.25 m height, both treatmentsf Gliricidia were at par and were better than Indigofera with orithout GFS. However, the canopy spread at top was significantlyigher in Gliricidia than with GFS. The treatments differed signifi-antly among each other with respect to the height at first branch.rovision of GFS lowered the height of first branch in Gliricidiaut increased in Indigofera. There was no significant differenceetween treatments with regard to number of branches. In general,FS did not have any species specific impact with respect to sur-ival, but showed differential impact in terms of parameters likelant height and height of first branch in the two hedge species.upplementation with GFS indicated a factor of competition in Gli-icidia, but favoured Indigofera performance.

Survival and growth performance of the grass species used aslter was satisfactory. Even after 30 months of planting, the averageurvival was 73% with clump diameter of 15.33 cm and number oflips increasing to 13.39. The difference among treatments was notound significant.

.2. Runoff and soil loss

Averaged over 3 years of observation, the efficacy ofndigofera + GFS was found superior (Table 2) with lowest runoff8.9%) and soil loss (5.0 Mg ha−1), followed by Gliricidia + GFS (10.7%unoff, 6.3 Mg ha−1 soil loss). A consistently lower runoff and soiloss under Indigofera + GFS was despite its poor growth perfor-

ance compared to Gliricidia (Table 1). It might be possible thathort and compact growth habit and better soil binding efficiencyf Indigofera roots resulted in lower runoff and soil loss, whenntegrated with GFS. On the other hand, performance of Indigofera

able 2unoff and soil loss under different treatments during 2002–2004.

Treatment Run off (%)

2002 2003 2004

Indigofera 12.35 13.01 12.9

Indigofera + GFS 8.1 9.45 9.1

Gliricidia 11.53 12.29 11.8

Gliricidia + GFS 10.35 11.27 10.5

Control 14.75 16.25 17.2

Sole GFS 12.75 14.16 14.5

.30 54.42 80.58 6.24

.69 7.21 0.87 NS

.74 13.91 10.24 16.21

alone (without GFS) was not satisfactory. This shows better com-plementary effect of Indigofera with GFS. A first line barrier of GFSmight have favoured better root growth of Indigofera that resultedin the best anti-erosive action. On the other hand, Gliricidia alonewas significantly better than Indigofera alone treatment indicatingthe necessity of GFS supplement for the best efficacy of Indigofera.The highest runoff and soil loss was observed under control fol-lowed by sole GFS treatment. In all treatments, addition of GFSsignificantly reduced both runoff and soil loss due to an additionalrow of barrier and the short growing local grass proving effectivein breaking the runoff velocity (Dass et al., 2011). Comparing thetreatments with the control showed the best runoff controllingtreatment (Indigofera + GFS) can reduce runoff by 45% and soil lossby about 48% and provision of only a row of GFS can reduce runoffby 14% and soil loss by about 23%. Gliricidia alone reduced runoff by26% and soil loss by 34% whereas GFS supplement further improvedthe conservation efficacy (runoff reduced by 33% and soil loss by35%). Even 1-year old hedgerows could reduce runoff by 63–71%and soil loss by 82–86% (Lin et al., 2009). In a comparative studyof grass species, Dass et al. (2011) showed double row of Sambutato be better with about 50% lower runoff and soil loss than othergrass species in sloping lands of eastern India. The grass barriers ofnapier (Pennisetum purpureum) and vetiver (Vetiveria zizanioides)reduced runoff by 54 and 12% and soil loss by 92 and 48% overcontrol, respectively (Owino and Gretzmacher, 2002).

3.3. Nutrient loss

As expected, nutrient loss was lowest under Indigofera + GFS dueto low runoff and soil loss. Under the particular treatment, loss ofSOC and available N, P and K were lower by 43, 56, 54 and 48%,respectively (Table 3). However, it was statistically at par with Gli-ricidia + GFS with regard to all the reported nutrients (Table 4). Allthe treatments were significantly better than control. The maxi-mum loss of nutrients was under control with annual loss of SOC,available N, P and K at 132.2, 7.35, 0.52 and 4.10 kg ha−1. Provi-sion of GFS reduced the loss of nutrients in all the treatments. For

instance, under Indigofera, additional GFS helped in reducing lossof SOC by 35%, available N by 45%, available P by 37% and available Kby 32%. The corresponding reduction in loss of nutrients due to GFSunder Gliricidia was 16, 18, 13 and 31% respectively. As compared

Soil loss (Mg ha−1)

Mean 2002 2003 2004 Mean

12.75 7.4 8.36 7.5 7.758.88 4.95 5.19 4.99 5.04

11.87 5.26 6.57 7.4 6.4110.71 5.9 7.1 6 6.3316.07 8.8 10.24 10.1 9.7113.80 6.5 7.85 8.1 7.48

36 N.K. Lenka et al. / Agriculture, Ecosystems and Environment 158 (2012) 31– 40

Table 3Loss of soil nutrients (kg ha−1 yr−1) under different treatments (mean of observations during 2002–2004).

Treatment Organic C Available N Available P Available K

Indigofera 116.25 5.84 0.38 3.13Indigofera + GFS 75.62 3.22 0.24 2.14Gliricidia 96.15 4.14 0.32 2.78

.41

.35

.21

ta

os0ltoeoSfftmflsNt

3

rtenwtuGrc1

TS

iN

Gliricidia + GFS 80.28 3Control 132.24 7Sole GFS 123.24 6

o control, sole GFS treatment could reduce the loss of SOC by 7%,vailable N by 16%, available P by 21% and available K by 17%.

The trend of results obtained is comparable with the findingsf Bhanavase et al. (2007) in rainfed regions of central India underunflower crop, where the loss of N, P and K was in the range of 3–6,.4–0.6 and 0.2–0.5 kg ha−1 yr−1. Contour hedgerows could reduce

oss of total N and total P by 81–85% and 91–93% respectively in arial under sloping conditions of China (Tao et al., 2012). The patternf nutrient loss, by and large, followed the trend of soil loss. Waterrosion affects surface horizon the most. Due to the low densityf SOC fractions than soil mineral particles and because of higherOC concentration in top soil, there is a preferential removal of SOCrom surface layers (Lal, 2005). Also, recurrent erosion obstructs theormation of stable soil–humus complex from the soil organic mat-er accumulated during the non-erosion period of the year. Thus,

ostly light fraction and particulate organic matter substances thatorm the major pool in the labile SOC fractions are expected to beost through the runoff water. In addition to the loss through soilediments and particulate and light fraction organic substances,

is also lost with water due to its high solubility as comparedo P.

.4. SOC concentration

The SOC was higher near the treatment bunds and graduallyeduced towards the plot. This trend was observed for all thereatments but was limited to only 0–15 cm and 15–30 cm soil lay-rs (Table 5). Effect of treatments and distance from bunds wereot observed in the 30–60 cm soil layer. Also with depth, thereas a decline in SOC under all the treatments. At 1 m bund dis-

ance, in the 0–15 cm soil layer the SOC was highest (5.90 g kg−1)nder Indigofera + GFS which was however statistically at par with

liricidia + GFS (5.82 g kg−1). The next best treatment was Gli-

icidia alone (4.91 g kg−1). The lowest SOC was observed underontrol which was significantly lower than the initial. In the5–30 cm soil layer, similar trend was observed among treatments

able 4tatistical significance level among paired treatments for runoff, soil loss and nutrient los

Paired treatments Runoff Soil loss Organic car

i–c * * *

igf–c * * *

g–c * * *

ggf–c * * *

gf–c * * *

i–igf * * *

i–g * * *

i–ggf * * *

i–gf * * *

igf–g * * *

igf–ggf * * NS

igf–gf * * *

g–ggf * * *

g–gf * * *

ggf–gf * * *

: Indigofera; igf: Indigofera + GFS; g: Gliricidia; ggf: Gliricidia + GFS; c: Control; gf: Sole GS – statistically not significant.* Significance at p < 0.05.

0.28 1.920.52 4.100.41 3.42

but significantly best treatment was proved to be Gliricidia + GFSfollowed by Indigofera + GFS. At 2 m bund distance, the SOC was sig-nificantly higher under Gliricidia + GFS followed by Indigofera + GFSin both 0–15 cm and 15–30 cm soil depths. The lowest SOC wasunder control followed by sole GFS treatment.

The difference among treatments was also apparent in thewhole plot SOC concentration. Gliricidia + GFS and Indigofera + GFSwere statistically at par and were significantly superior to othertreatments in both 0–15 cm and 15–30 cm soil depths. Gliricidiaalone was next best and was significantly better than Indigofera andGFS treatments. Lowest SOC was observed under control in both0–15 cm (3.84 g kg−1) and 15–30 cm (3.82 g kg−1) soil layers. Reha-bilitation studies involving land use changes, many times, show thetrend of increase in SOC concentration after 4–5 years dependingupon soil type and climate. For instance, agricultural lands whenput to poplar plantations were a source of C in the initial 2 years butafterwards the SOC content increased (Arevalo et al., 2011). A 15%increase in SOC content in the surface soil was observed in a 12-yearhedgerow experiment with Gliricidia sepium and Leucaena leuco-cephala on a Nigerian Alfisol (Kang et al., 1999). Even after 5 yearsof trial with Inga edulis, 12% increase in SOC content was observedunder a Peruvian Ultisol (Alegre and Rao, 1996). The trend of SOCincrease may be higher if the degraded lands are put under agro-forestry practices. For instance, Lenka et al. (2012) reported 89%increase in SOC of a degraded land ravaged by shifting cultivation,after 6 years under a horti-silvi-pastural system.

With GFS supplement, increase in SOC was observed for all thetreatments at 1 m and 2 m bund distance and for whole plot inthe 0–15 cm and 15–30 cm soil layers. The highest improvement(31% at 1 m bund distance and 6% for whole plot) was observedin Indigofera which indicated better performance of the specieswith additional GFS. This might be due to the slow initial growth of

Indigofera as compared to Gliricidia. The GFS supplement providedprotection from erosion in the initial establishment phase whichhelped retaining soil and nutrients and also improving the growthand establishment of the hedge species.

s parameters, tested at p < 0.05.

bon Available N Available P Available K

* * *

* * *

* * *

* * *

* * *

* * *

* NS *

* * *

* * NS* * *

NS NS NS* * *

* * *

* * *

* * *

FS.

N.K. Lenka et al. / Agriculture, Ecosystems and Environment 158 (2012) 31– 40 37

Table 5Soil organic carbon (g kg−1) at 1 m and 2 m bund distance and for the whole plot for different soil layers under selected treatments after 5 years of study.

Treatment SOC (g kg−1)

1 m 2 m Plot

0–15 cm 15–30 cm 30–60 cm 0–15 cm 15–30 cm 30–60 cm 0–15 cm 15–30 cm 30–60 cm

Indigofera 4.52 4.30 3.42 4.22 4.13 3.38 4.16 3.94 3.40Indigofera + GFS 5.90 5.24 3.40 4.90 4.45 3.41 4.52 4.08 3.37Gliricidia 4.91 4.90 3.55 4.70 4.41 3.44 4.30 4.02 3.40Gliricidia + GFS 5.82 5.41 3.37 5.05 4.52 3.52 4.50 4.10 3.43Control 3.96 3.75 3.42 3.82 3.74 3.46 3.84 3.64 3.40Sole GFS 4.32 4.15 3.47 4.18 4.02 3.39 4.10 3.82 3.38Initial 4.08 3.85 3.44 4.04 3.94 3.40 3.96 3.83 3.40LSD (0.05) 0.11 0.16 NS 0.14 0.18 NS 0.10 0.16 NS

G

3

psopsewIwafbGGiug1

r(rstSrgsw

TS

G

CV, % 2.64 3.68 3.24 3.86

FS – Grass filter strip.

.5. Soil organic carbon stock (soil layer wise)

The SOC stock was significant for treatments at all the sam-ling points, but the trend was limited to 0–15 cm and 15–30 cmoil depths only (Table 6). Below 30 cm, no significant effect wasbserved. The SOC stock followed the trend of 1 m > 2 m > wholelot. At 1 m bund distance in the 0–15 cm soil layer, highest SOCtock of 13.14 Mg ha−1 was observed with GFS supplement inither of the hedge species and both being statistically at parith each other. Gliricidia alone was the next best followed by

ndigofera alone and sole GFS. The SOC stock in the control plotas significantly lower by 0.25 Mg ha−1 than the initial indicating

gradual decline in SOC in unprotected lands. The pattern of dif-erence among treatments was similar in the 15–30 cm soil layerut Gliricidia + GFS proved superior followed by Indigofera + GFS,liricidia, Indigofera and the lowest under control. Provision ofFS benefited all the treatments including control and the gain

n SOC stock ranged from 0.76 Mg ha−1 in control to 2.86 Mg ha−1

nder Indigofera in the 0–15 cm soil layer and the correspondingain in the 15–30 cm soil layer was 0.76 Mg ha−1 under control to.94 Mg ha−1 in Indigofera.

At 2 m bund distance, the SOC stock was highest under Gli-icidia + GFS, both in the 0–15 cm (11.51 Mg ha−1) and 15–30 cm10.44 Mg ha−1) soil layers. This treatment was significantly supe-ior to Indigofera + GFS in both the soil depths. The lowest SOCtock was observed under control followed by GFS only. The unpro-ected control plot showed 0.45 Mg ha−1 and 0.36 Mg ha−1 lowerOC stock than the initial in the 0–15 cm and 15–30 cm soil layers,espectively. Provision of GFS also benefitted the SOC stock and the

ain was highest (1.35 Mg ha−1) under Indigofera in the 0–15 cmoil layer and the corresponding gain in the 15–30 cm soil layeras 0.49 Mg ha−1.

able 6oil organic carbon stock (Mg ha−1) at 1 m and 2 m bund distance and for the whole plot f

Treatment SOC stock (Mg ha−1)

1 m 2 m

0–15 cm 15–30 cm 30–60 cm 0–15 cm

Indigofera 10.24 9.93 8.21 9.68

Indigofera + GFS 13.10 11.87 8.06 11.03

Gliricidia 11.19 11.25 8.57 10.86

Gliricidia + GFS 13.14 12.25 8.04 11.51

Control 9.15 8.89 8.26 8.94

Sole GFS 9.91 9.65 8.33 9.66

Initial 9.36 8.95 8.31 9.39

LSD (0.05) 0.18 0.34 NS 0.36

CV, % 1.82 3.63 2.78 3.84

FS – Grass filter strip.

4.76 3.20 2.82 4.52 3.24

The SOC stock of the whole plot was relatively lower thanthat of sampling points at 1 m and 2 m bund distance. This wasprimarily because the under-tree effects as well as the erosionresisting effects of hedges were better pronounced near the bunds(Walter et al., 2003). In addition to litter fall and root biomassturnover, a favorable soil moisture and temperature regime mighthave enabled better microbial activity and SOC profile under thehedges. Most of the nutrients and fine soil particles that are washedwith runoff water also get deposited due to barrier effect of hedges(Tao et al., 2012). Gliricidia + GFS and Indigofera + GFS were at parwith each other and were significantly better than other treatmentsin 0–15 cm and 15–30 cm soil depths. As compared to the initial,the SOC stock in the best treatment was higher by 1.19 Mg ha−1

in the 0–15 cm soil layer and by 0.53 Mg ha−1 in the 15–30 cmsoil layer. This is equivalent to sequestration of 0.238 Mg ha−1 yr−1

of SOC in the 0–15 cm soil layer and 0.106 Mg ha−1 yr−1 in the15–30 cm soil layer. The next best treatment was Gliricidia thatwas significantly superior to Indigofera in both the soil depths.As compared to the initial, there was a decline of 0.22 Mg ha−1

of SOC stock in the 0–15 cm and 15–30 cm soil layers, indicat-ing a loss of about 88 kg ha−1 yr−1 of SOC from the top 30 cmsoil.

The trend of results compares well with the findings of otherresearchers in similar climatic conditions. An increase in SOC stockin the surface soil by 0.23–2.38 Mg ha−1 due to hedgerow inter-cropping has been reported in degraded soils in the tropics (Alegreand Rao, 1996; Kang et al., 1999). Hedgerows planted in a series cancontribute to 13–38% increase in SOC storage (Walter et al., 2003).The difference in SOC accrual is a function of time as well as soil

type. Under restorative practices, rate of SOC increment is lower incoarse textured soils than fine textured soils (Albrecht and Kandji,2003) which may be due to formation of clay–SOM complexes.

or different soil layers under selected treatments after 5 years of study.

Plot

15–30 cm 30–60 cm 0–15 cm 15–30 cm 30–60 cm

9.66 8.21 9.61 9.22 8.1610.15 8.29 10.31 9.49 8.0910.45 8.41 9.93 9.53 8.1610.44 8.40 10.40 9.47 8.13

8.86 8.46 8.99 8.74 8.219.41 8.14 9.53 8.94 8.119.22 8.21 9.21 8.96 8.210.28 NS 0.19 0.20 NS3.08 2.82 2.36 2.65 2.63

38 N.K. Lenka et al. / Agriculture, Ecosystems

Table 7Soil organic carbon stock (Mg ha−1) in the 0.60 m soil profile at 1 m and 2 m bunddistance and for the whole plot under different treatments after 5 years of study.

Treatment SOC stock in 0.60 m profile

1 m 2 m plot

Indigofera 28.38 27.55 26.99Indigofera + GFS 33.03 29.47 27.89Gliricidia 31.01 29.72 27.62Gliricidia + GFS 33.39 30.35 28.00Control 26.30 26.26 25.94Sole GFS 27.89 27.21 26.58Initial 26.62 26.82 26.38LSD (0.05) 0.30 0.34 0.34

G

3

iatoGast1Is(aeatst

3

rIdb1HtnG

TSa

G

CV, % 1.56 1.65 1.42

FS – Grass filter strip.

.6. SOC stock in 0.6 m soil profile

The SOC stock computed for the 0.6 m profile (Table 7) wasn the range of 28.0 Mg ha−1 for the whole plot to 33.39 Mg ha−1

t 1 m bund distance in the best treatment (Gliricidia + GFS). Thereatments in terms of SOC stock were by and large in the orderf Gliricidia + GFS > Indigofera + GFS > Gliricidia > Indigofera > SoleFS > Initial > Control, for the three sampling points (1 m, 2 mnd whole plot). Provision of GFS contributed to a gain in SOCtock and the gain ranged from 0.38 Mg ha−1 under Gliricidiao 1.0 Mg ha−1 under Indigofera in the whole plot samples. At

m bund distance, the gain was higher with 4.65 Mg ha−1 underndigofera to 1.59 Mg ha−1 under control. The improvement in SOCtock in the best treatment ranged from 7.03 (1 m) to 2.06 Mg ha−1

plot) as compared to control and 6.77 (1 m) to 1.62 Mg ha−1 (plot)s compared to the initial. The loss in SOC stock due to continuedrosion as revealed from the lowered SOC stock in the controls compared to the initial was to the tune of 0.44 Mg ha−1 forhe whole plot, 0.32 Mg ha−1 for 1 m and 0.56 Mg ha−1 for 2 mamples. Thus, there was a loss of SOC by 60–112 kg ha−1 yr−1 dueo continued erosion.

.7. SOC build up rate and SOC sequestration rate

The SOC build up rate was significant for the treatments and theate under all treatments was higher than under control (Table 8).n general for all treatments, the rate was higher at 1 m bundistance and gradually reduced towards the plot. The highest SOCuild up rate was observed under Gliricidia + GFS, ranging from.354 Mg ha−1 yr−1 at 1 m to 0.352 Mg ha−1 yr−1 for the whole plot.

owever, this treatment was significantly better than the next best

reatment (Indigofera + GFS) at 1 m and 2 m bund distance only,ot for plot samples. In general, treatments were in the order ofliricidia + GFS > Indigofera + GFS > Gliricidia > Indigofera > Sole GFS.

able 8oil organic carbon build up rate (Mg ha−1 yr−1) and SOC sequestration rate (Mg ha−1 yr−

fter 5 years of study.

Treatment SOC build up rate (Mg ha−1 yr−1)

1 m 2 m

Indigofera 0.352 0.146

Indigofera + GFS 1.282 0.53

Gliricidia 0.878 0.58

Gliricidia + GFS 1.354 0.706

Control −0.064 −0.112

Sole GFS 0.254 0.078

Initial – –

LSD (0.05) 0.06 0.12

CV, % 9.54 29.65

FS–Grass filter strip.

and Environment 158 (2012) 31– 40

Supplemental row of GFS increased the SOC build up rate signifi-cantly under all the hedge species and control. Under Indigofera,the gain was the maximum and ranged from 0.93 Mg ha−1 yr−1

at 1 m bund distance to 0.18 Mg ha−1 yr−1 for the whole plot. Thecorresponding gain was 0.476–0.076 Mg ha−1 yr−1 in Gliricidiaand 0.318–0.128 Mg ha−1 yr−1 in control, at 1 m bund distanceand for plot, respectively. A negative value of SOC build up rateunder control showed a net loss of 0.06–0.112 Mg ha−1 yr−1 of SOC,equivalent to 60–112 kg ha−1 yr−1 of SOC under unprotected con-ditions. On the other hand, the best conservation treatment couldresult in building SOC by 352 kg ha−1 yr−1 to 1354 kg ha−1 yr−1.

The SOC sequestration rate computed with respect to controlwas higher in all treatments than the corresponding values of SOCbuild up rate (Table 8). This was highest in the Gliricidia + GFStreatment ranging from 1.418 Mg ha−1 yr−1 at 1 m bund distance to0.412 Mg ha−1 yr1 yr−1 for the plot. The order of difference amongtreatments followed similar trend like that of SOC build up rate.Provision of GFS resulted gain in SOC sequestration rate under allthe hedge treatments and also in control. The gain under Indigoferawas maximum and varied from 0.93 Mg ha−1 yr−1 at 1 m bund dis-tance to 0.18 Mg ha−1 yr−1 for the plot. The corresponding gain was0.476–0.076 Mg ha−1 yr−1 under Gliricidia at 1 m bund distance andfor plot, respectively. The SOC sequestration rate was higher thanthe SOC build up rate because, SOC build up rate when comparedwith initial, does not take into account the factor of on-going ero-sion.

Though there are few studies on the sequestration potential ofSOC due to soil erosion control in the study region, but the find-ings may be in agreement with the estimates of 0.5 Mg ha−1 yr−1

for agricultural lands of tropical regions (Lal, 2008). Under arablefarming, conversion from plough till to no-till was reported tosequester at a mean rate of 570 ± 140 kg C ha−1 yr−1 (West andPost, 2002). Pacala and Socolow (2004) estimated a sequestrationpotential of 0.5–1.0 Pg C yr−1 by 2050 with conversion of ploughtillage to no-till farming on 1600 M ha of croplands along withadoption of conservation-effective measures. Higher SOC buildup is possible with complete land cover change with pasturesor agroforestry systems due to increased rate of organic matteraddition and retention (Sanchez, 2000; Sharrow and Ismail, 2004;Swami and Puri, 2005; Lenka et al., 2012). A SOC build up rate of3.5–4.5 Mg ha−1 yr−1 could be possible with Stylosanthes and grasscover in degraded hillock sites as reported by Lenka et al. (2012).If lands are degraded, the response to restorative measures may behigher and thus the C sequestration rate may be higher in the initialyears before reaching a plateau, as compared to crop fields culti-vated with management practices. For instance, the rate of change

in SOC stock observed after 21 years in a rice–lentil cropping sys-tem, varied from 0.043 to 0.462 Mg ha−1 yr−1 (Srinivasarao et al.,2011), which is relatively lower as compared to the findings of thisstudy.

1) at 1 m and 2 m bund distance and for the whole plot under different treatments

SOC sequestration rate (Mg ha−1 yr−1)

Plot 1 m 2 m Plot

0.150 0.416 0.258 0.210.33 1.346 0.642 0.390.276 0.942 0.692 0.3360.352 1.418 0.818 0.412

−0.06 – – –0.068 0.318 0.19 0.128– 0.064 0.112 0.0880.08 0.07 0.12 0.06

32.10 8.36 24.41 21.05

N.K. Lenka et al. / Agriculture, Ecosystems and Environment 158 (2012) 31– 40 39

i :Indigofera, igf : Indigofera + GFS, g: Gliricidia, ggf: Gliricidia + GFS, c: Control, gf: GFS

LSD : Least significant difference, at P < 0.05

0

20

40

60

80

100

120

140

160

180

i igf g ggf c gf LSD

Treatments

34 mm, 12 days

(mm

)

0

20

40

60

80

100

120

140

160

180

i igf g ggf c gf LSD

1 m

2 m

Centre

Treatments

(mm

)

47 mm, 17 days(b)

(a)

Ffe

eetiobneSt(

cooc1esl2

3

scsui

Table 9Yield of test crop (finger millet) under different treatments.

Treatment Yield (kg ha−1)

2004 2005 Mean

Indigofera 1080.0 1352.2 1216.1Indigofera + GFS 1263.1 1471.1 1367.1Gliricidia 1227.8 1225.6 1226.7Gliricidia + GFS 1404.7 1421.1 1412.9Control 874.1 1030.0 952.05Sole GFS 1057.8 1083.3 1070.5LSD (0.05) 43.1 83.6 –

ig. 3. Volumetric soil moisture storage (�, mm) in the 0.6 m soil profile under dif-erent treatments at (a) 34 mm evaporativity and 12 days dry spell and (b) 47 mmvaporativity and 17 days dry spell.

In fact, erosion control and SOC status are complementary toach other. A higher SOC status permits less erosion where as betterrosion control promotes SOC build up and improves soil aggrega-ion and soil quality. The increase in SOC build up rate as observedn this study may be due to the effect of conservation treatmentsn soil reclamation (Albrecht and Kandji, 2003) and also due to thearrier effect and the consequent retention of sediments, SOC andutrients in treated plots (Walter et al., 2003). The ameliorationffect of erosion control measures primarily arise from the higherOC and the role of SOC in promoting soil aggregation, which inurn, conserves soil organic matter and enhances the SOC functionsCarter, 2002).

An estimate to segregate the effect of the two mechanismsan be obtained from the difference in the SOC build up ratef the whole plot and that at 1 m bund distance. Thus, in theverall best treatment (Gliricidia + GFS), 0.352 Mg ha−1 yr−1 of SOCould be sequestered due to soil reclamation whereas about.002 Mg ha−1 yr−1 of SOC was primarily retained due to the barrierffect and contributing to the higher SOC build up. A higher nutrienttatus particularly N in the conservation treatments, due to reducedoss through runoff, also contributed to higher C sequestration (Lal,008).

.8. Soil moisture

The volumetric soil moisture content (�) computed up to 0.6 moil profile was higher at 1 m bund distance and reduced towards

entre of the plots (Fig. 3). At 1 m bund distance, � in the 0.6 moil profile ranged from 164 mm under Gliricidia + GFS to 127 mmnder control, at a cumulative evaporation of 34 mm correspond-

ng to 12 days of dry period. Gliricidia + GFS was significantly better

CV, % 5.84 8.45 –

GFS – Grass filter strip.

than Indigofera + GFS (156 mm). All the treatments with GFS weresignificantly better than the corresponding no-GFS treatment. Sim-ilar trend continued at 2 m bund distance, but � was lower. At thecentre of the plot, � was the least as expected and varied from 114to 142 mm with the highest storage under Gliricidia + GFS. Highersoil moisture storage near the treatment bunds was due to the bar-rier effect resulting in deposition of nutrients and also due to thehigher SOC built up from the litter fall (Walter et al., 2003; Dasset al., 2011). Competition between crops and vegetative barriersfor water is an important factor while selecting species for use inerosion control. In a similar study in Central Kenya soil moistureregime was 56–77% better near Leucaena barriers and though Leu-caena was observed to have a complementary water use pattern,Napier grass depleted the available water status (Guto et al., 2011).

With increased dry period and at an evaporativity of 47 mmcorresponding to a dry period of 17 days, � reduced by 9–15 mmat 1 m bund distance, 8–16 mm at 2 m and 7–14 mm at centre ofthe plot. The trend of difference among treatments was similarto evaporativity of 34 mm. At all sampling points, Gliricidia + GFSretained higher soil moisture followed by Indigofera + GFS and thelowest was in control. The variation in � was from 112–155 mm at1 m, 116–138 mm at 2 m and 102–132 mm at centre of the plot. Ingeneral, provision of GFS resulted higher moisture storage undereach treatment. The gain in � due to GFS supplement ranged from9 to 12 mm and 6–15 mm at evaporativity of 34 mm and 47 mm,respectively.

3.9. Yield of test crop

The yield data showed a similar trend of difference amongtreatments with highest crop yield under Gliricidia + GFS, whichwas at par with Indigofera + GFS (Table 9). In 2004 crop season,yield varied from 874.1 kg ha−1 in control to 1404.7 kg ha−1 in Gli-ricidia + GFS. The yield levels in 2005 ranged from 1030 kg ha−1

in control to 1471.1 kg ha−1 in Indigofera + GFS. However, Gliri-cidia + GFS and Indigofera + GFS were at par with each other. Thegeneral trend of difference among treatments was in the orderof Gliricidia + GFS ≈ Indigofera + GFS > Gliricidia ≈ Indigofera > SoleGFS > Control. Provision of GFS contributed to increase in crop yieldranging from 177 to 183 kg ha−1 in 2004 and 53–195 kg ha−1 in2005 crop season, corresponding to 15–22% increase in 2004 and5–16% increase in 2005. As compared to control, the best treatmentcould produce 61% and 43% higher crop yield which was primarilydue to better SOC status and higher soil moisture storage (Sudhishriet al., 2008). Similar increase in maize yield was observed with Leu-caena barrier and the better growth performance at areas awayfrom barriers in the terraces led to 14% higher yield and com-pensated the loss of yield near the barriers due to crop–grass

competition (Guto et al., 2011). The adverse effect of soil moisturedepletion due to higher water use by grass barriers such as Napier(Guto et al., 2011) was not observed in this study because the grassbarrier provided was supplemental in a single row and with short

4 stems

gbbesr1

4

staip0tmocatcwistIst

A

AnrSa(

R

A

A

A

B

B

B

B

CC

C

CD

0 N.K. Lenka et al. / Agriculture, Ecosy

rowth habit. A reduction in crop yield up to 1.08 m from the barrierunds was observed with hill broom grass (Thysanolaena maxima)arriers and up to 0.9 m with Saccharum spp. barriers (Sudhishrit al., 2008). But in this experiment, both the leguminous hedgepecies providing N-rich litter fall and residue biomass turnoveresulted a better soil moisture regime that was even higher after7 days of dry spell.

. Conclusion

The study showed that integrations of hedgerow and grass filtertrips are more effective than the sole ones for soil erosion con-rol, carbon sequestration and productivity increase in the slopinggricultural lands of eastern India. The hedge species of Gliricidiantegrated with GFS was the best option in terms of most of thearameters studied, which also sequestered SOC at the rate of.412–1.418 Mg ha−1 yr−1. Higher soil moisture storage increasedhe grain yield of finger millet by 49% under Gliricidia + GFS treat-

ent. Supplemental row of GFS significantly reduced the lossesf water runoff, soil and nutrients in all the treatments includingontrol, and led to higher SOC sequestration and soil moisture stor-ge. As compared to pre-treated initial, decrease of SOC stock inhe control indicated a significant impact of on-going erosion pro-ess in the unprotected lands. The best treatment of Gliricidia + GFSas followed by Indigofera + GFS in most of the parameters stud-

ed, though the latter showed satisfactory performance only withupplemental GFS. The results recommend the use of grass fil-ers additionally to the hedge species (particularly required forndigofera) for the complementary benefit of both hedge and grasspecies and for maintaining better soil quality in sloping agricul-ural lands of eastern India.

cknowledgements

Authors acknowledge the World Bank and Indian Council ofgricultural Research (ICAR) sponsored National Agricultural Tech-ology Project (NATP) for funding this participatory on-farmesearch project. The encouragement and guidance of Dr. V.N.harda, Director, Central Soil and Water Conservation Researchnd Training Institute, Dehradun and Dr. B. Venkateswarlu, AEDRainfed), CRIDA, Hyderabad are thankfully acknowledged.

eferences

lbrecht, A., Kandji, S.T., 2003. Carbon sequestration in tropical agroforestry systems.Agric. Ecosyst. Environ. 99, 15–27.

legre, J.C., Rao, M.R., 1996. Soil and water conservation by contour hedging in thehumid tropics of Peru. Agric. Ecosyst. Environ. 57, 17–25.

revalo, C.B.M., Bhatti, J.S., Chang, S.X., Sidders, D., 2011. Land use change effects onecosystem carbon balance: from agricultural to hybrid poplar plantation. Agric.Ecosyst. Environ. 141, 342–349.

aker, J.M., Ochsner, T.E., Venterea, R.T., Griffis, T.J., 2007. Tillage and soil carbonsequestration – what do we really know. Agric. Ecosyst. Environ. 118, 1–5.

hanavase, D.B., Deshpande, A.N., Pawar, A.B., 2007. Effect of vegetative barriers onresource conservation and productivity of sunflower in Inceptisols of Maharash-tra. Indian J. Soil Conserv. 35, 238–241.

lack, C.A., 1965. Methods of Soil Analysis, Part 1. American Society of Agronomy,Madison, Wisconsin.

lanco-Canqui, H., Schlegel, A.J., Heer, W.F., 2011. Soil-profile distribution of carbonand associated properties in no-till along a precipitation gradient in the centralGreat Plains. Agric. Ecosyst. Environ. 144, 107–116.

arter, M.R., 2002. Soil quality for sustainable land management. Agron. J. 94, 38–47.hunale, G.L., 2004. Evaluation of different grass species for soil binding and soil

aggregation properties under sub-montane zone of Maharashtra. Indian J. SoilConserv. 32, 24–27.

hung, H., Ngo, K.J., Plante, A.F., Six, J., 2010. Evidence for carbon saturation in ahighly structured and organic matter rich soil. Soil Sci. Soc. Am. J. 74, 130–138.

ochran, W.G., Cox, G.M., 1963. Experimental Designs. Wiley Inc., New York.ass, A., Sudhishri, S., Lenka, N.K., Patnaik, U.S., 2011. Runoff capture through vege-

tative barriers and planting methodologies to reduce erosion and improve soil

and Environment 158 (2012) 31– 40

moisture, fertility and crop productivity in southern Orissa, India. Nutr. Cycl.Agroecosyst. 89, 45–57.

Eswaran, H., Lal, R., Reich, P.F., 2001. Land degradation: an overview. In: Bridges, E.M.,Hannam, I.D., Oldeman, L.R., Pening de Vries, F.W.T., Scherr, S.J., Sompatpanit,S. (Eds.), Responses to Land Degradation. Proc. 2nd International Conference onLand Degradation and Desertification, Khon Kaen, Thailand. The Oxford Press,New Delhi.

Guto, S.N., de Ridder, N., Giller, K.E., Pypers, P., Vanlauwe, B., 2011. Minimum tillageand vegetative barrier effects on crop yields in relation to soil water content inthe Central Kenya highlands. Field Crops Res. 132, 129–138.

Jackson, M.L., 1973. Soil Chemical Analysis. Prentice Hall of India Pvt. Ltd., New Delhi.Kang, B.T., Caveness, F.E., Tian, G., Kolawole, G.O., 1999. Long-term alley cropping

with four species on an Alfisol in southwest Nigeria – effect on crop performancesoil chemical properties and nematode population. Nutr. Cycl. Agroecosyst. 54,145–155.

Kumar, P., 2011. Capacity constraints in operationalisation of payment for ecosystemservices (PES) in India: evidence from land degradation. Land Degrad. Dev. 22,432–443.

Lal, R., 2005. Soil erosion and carbon dynamics. Soil Till. Res. 81, 137–142.Lal, R., 2008. Carbon sequestration. Philos. Trans. R. Soc. B 363, 815–830.Lal, R., Follett, R.F., Kimble, J., Cole, C.V., 1999. Managing US cropland to sequester

carbon in soil. J. Soil Water Conserv. 54, 374–381.Lenka, N.K., Choudhury, P.R., Sudhishri, S., Dass, A., Patnaik, U.S., 2012. Soil aggre-

gation carbon build up and root zone soil moisture in degraded sloping landsunder selected agroforestry based rehabilitation systems in eastern India. Agric.Ecosyst. Environ. 150, 54–62.

Lin, C., Tu, S., Huang, J., Chen, Y., 2009. The effect of plant hedgerows on the spatialdistribution of soil erosion and soil fertility on sloping farmland in the purple-soilarea of China. Soil Till. Res. 105, 307–312.

Ogle, S.M., Swan, A., Paustian, K., 2012. No-till management impacts on crop produc-tivity carbon input and soil carbon sequestration. Agric. Ecosyst. Environ. 149,37–49.

Owino, J., Gretzmacher, R., 2002. Performance of narrow strips of vetiver grassand napier grass as barriers against runoff and soil loss on a clay loam soil inKenya. In: Proceedings of the Conference on International Agricultural ResearchDevelopment held at Deutscher, Witzenhausen during October 9–11, 2002.

Pacala, S., Socolow, R., 2004. Stabilization wedges: solving the climate problem forthe next 50 years with current technologies. Science 305, 968–972.

Planning Commission, Government of India, 2012. Percentage of pop-ulation below poverty line by States/UTs (1973–74 to 2004–05),http://planningcommission.gov.in.

Saha, S., Nair, P.K.R., Nair, V.D., Kumar, B.M., 2010. Carbon storage in relation to soilsize-fractions under some tropical tree-based land-use systems. Plant Soil 328,433–446.

Sanchez, P.A., 2000. Linking climate change research with food security and povertyreduction in the tropics. Agric. Ecosyst. Environ. 82, 371–383.

Schlesinger, W.H., 2000. Carbon sequestration in soils: some cautions amidst opti-mism. Agric. Ecosyst. Environ. 82, 121–127.

Sharma, S.C., Mann, J.S., Mehta, R.S., 2002. Performance of vegetative barriers inestablishment of Cenchrus pastures in sandy loam soils of semiarid regions. In:Dhyani, S.K., Tripathy, K.P., Singh, R., Raizada, A., Sharma, N.K., Mishra, A.S., Shri-mali, S.S., Dhyani, B.L., Sharma, A.R., Khola, O.P.S. (Eds.), Resource Conservationand Watershed Management: Technology Options and Future Strategies. IndianAssociation of Soil and Water Conservationists, CSWCRTI, Dehradun.

Sharrow, S.H., Ismail, S., 2004. Carbon and nitrogen storage in western Oregon agro-forests, forests, and pastures. Agroforest. Syst. 60, 123–130.

Sikka, A.K., Singh, D.V., Gadekar, H., Madhu, M., 2000. Diagnostic Survey and Zonationof Eastern Ghats for Natural Resource Conservation and Management. CentralSoil and Water Conservation Research and Training Institute, Research Center,Ootacamund, India, p. 82.

Srinivasarao, C., Venkateswarlu, B., Lal, R., Singh, A.K., Vittal, K.P.R., Kundu, S., Singh,S.R., Singh, S.P., 2011. Long-term effects of soil fertility management on carbonsequestration in a rice–lentil cropping system of the Indo-Gangetic Plains. SoilSci. Soc. Am. J. 76, 168–178.

Srivastava, R.C., Singhandhupe, R.B., Mohanty, R.K., 2004. Integrated farmingapproach for runoff recycling systems in humid plateau areas of eastern India.Agric. Water Manage. 64, 197–212.

State of Environment Report - India, 2009. Ministry of Environment and Forests,Govt. of India, New Delhi.

Sudhishri, S., Dass, A., Lenka, N.K., 2008. Efficacy of vegetative barriers for rehabili-tation of degraded hill slopes in eastern India. Soil Till. Res. 99, 98–107.

Swami, S.L., Puri, S., 2005. Biomass production and C-sequestration of Gmelinaarborea in plantation and agroforestry system in India. Agroforest. Syst. 64,181–190.

Tao, W., Bo, Z., Lizhong, X., 2012. Effects of contour hedgerow intercropping onnutrient losses from sloping farmland in three Gorges area, China. J. Mt. Sci.9, 105–114.

Walter, C., Merot, P., Layer, B., Dutin, G., 2003. The effect of hedgerows on soil organiccarbon storage in hillslopes. Soil Use Manage. 19, 201–207.

West, T.O., Post, W.M., 2002. Soil organic carbon sequestration rates by tillage andcrop rotation: a global data analysis. Soil Sci. Soc. Am. J. 66, 1930–1946.

Xu, Q.X., Wang, T.W., Cai, C.F., Li, Z.X., Shi, Z.H., 2012. Effects of soil conservation onsoil properties of citrus orchards in the three Gorges area, China. Land Degrad.Dev. 23, 34–42.