Ig

S

T

a

ARRA

KSCRSE

1

ssEoc2a

UK

A

h0

Ecological Engineering 71 (2014) 458–465

Contents lists available at ScienceDirect

Ecological Engineering

journa l h om epa ge: www.elsev ier .com/ locate /eco leng

mproved soil fertility from compost amendment increases rootrowth and reinforcement of surface soil on slopes

uzanne Donn, Ron E. Wheatley, Blair M. McKenzie, Kenneth W. Loades, Paul D. Hallett ∗,1

he James Hutton Institute, Invergowrie, Dundee, United Kingdom

r t i c l e i n f o

rticle history:eceived 17 September 2013eceived in revised form 19 June 2014ccepted 20 July 2014

eywords:lopeompostoot reinforcementoil physicscoengineering

a b s t r a c t

Vegetation such as grasses and shrubs can improve slope stability, reducing the risk of shallow failuresonce roots have permeated soil to enhance cohesion. However, establishment of vegetation is hinderedby poor soil fertility, frequently a characteristic of disturbed soils used in engineering projects. We eval-uated whether compost could improve the rate of vegetation establishment and hence soil mechanicalreinforcement by plant roots and therefore protect against shallow failures. Over 1200 t of material wasformed into a slope 40 m long × 15 m wide, with an experimental soil slope angle of 20◦. Washings fromrecycled mineral fill were used for the surface soil. Five amendment treatments were replicated threetimes in strips of 1.5 m by 8 m in a randomised block design on this slope; treatments were a no com-post control, standard compost addition at a rate of 35 t ha−1 and a high level compost amendment at300 t ha−1, applied either to the surface or incorporated into the topsoil to 10 cm.

Thirteen weeks after planting an amenity grass mix, vegetation cover increased up to 6 times comparedto the control for 35 t ha−1 surface applied compost and similarly 20 times for 300 t ha−1 compost thathad been surface applied. Root length density was about 3 km m−3 with no added compost and about30 km m−3 for 300 t ha−1 added compost. At 35 t ha−1 compost, peak shear stress of the vegetated soil at5 cm depth was not affected, but it almost doubled with 300 t ha−1 compost compared to no amendment.

−1

Cohesion from plant roots was 8.1 kPa for 300 t ha , in comparison to 2.1 kPa for no amendment and2.3 kPa for 35 t ha−1 compost. Whereas surface application resulted in better vegetation cover, there wereno differences in peak shear stress between plots with surface application or incorporation of compost.This study provided experimental evidence in the field that compost improvement to soil fertility has apositive impact on soil stabilisation by plant roots.

2tvbgip

. Introduction

Composts and other organic residues are increasingly used inlope stabilisation projects and agriculture as a mulch to provideurface protection against soil erosion (Bazzoffi et al., 1998;dwards et al., 2000; Hansen et al., 2009). It is well known thatrganic residues can improve plant performance, as shown in agri-

ultural (Warman et al., 2009), agroforestry (Gruenewald et al.,007), soil restoration (Ohsowski et al., 2012; Vaidya et al., 2008)nd engineered slope studies (Faucette et al., 2009; Persyn et al.,

∗ Corresponding author at: Institute of Biological and Environmental Sciences,niversity of Aberdeen, Cruickshank Building, St Machar Drive, AB24 3UU, Unitedingdom. Tel.: +44 01224 272264; fax: +44 01224 272703.

E-mail address: paul.hallett@abdn.ac.uk (P.D. Hallett).1 Present address: CSIRO Plant Industry, GPO Box 1600, Canberra 2601, ACT,ustralia.

Plos

st2aa

ttp://dx.doi.org/10.1016/j.ecoleng.2014.07.066925-8574/© 2014 Elsevier B.V. All rights reserved.

© 2014 Elsevier B.V. All rights reserved.

004). To our knowledge, however, no systematic study has inves-igated the combined impact of composts and potential improvedegetation growth on the mechanical reinforcement of soil slopesy plant roots. Such findings would be useful in developing newreen engineering solutions for slope stability, where potentialmproved plant performance and the use of a recycled material willroduce multiple environmental benefits (Mickovski et al., 2013).lant roots bind the soil across potential slip planes so that at shal-ow depths where root density is abundant there is a decreased riskf shallow soil failures. Roots act as natural anchors resulting in amaller carbon footprint from engineering projects.

The relative impact of organic residues on plant growth andoil stability tends to be greater for soils with poor initial fer-

ility (Diacono and Montemurro, 2010; Fernandez-Getino et al.,012). In engineering projects, physical disturbance of soils thatre excavated, stockpiled and then used to cover embankmentsnd cuttings can decrease fertility, thereby hindering the success

ngine

opttKctppo

c2icctmmeefoiaata(osctt

welw2ist

im3aTdtticsact

2

2

C

(tt

beoRsltwmtptg

twahfmiE8pmcsdcRcFa

2

eAsoaaw2

pesswawt2p

S. Donn et al. / Ecological E

f revegetation (Cao et al., 2010). In urban or mine reclamationrojects, soils can be extremely infertile so revegetation is limitedo hardy species that may fix nitrogen or requires interventiono improve soil conditions (Boyer and Wratten, 2010; Mallik andarim, 2008; Montoro et al., 2000). Recycled soils formed fromrushed building waste, or mine spoils that have low concentra-ions of organic carbon, are also used as fill in some engineeringrojects (Marando et al., 2011; Said-Pullicino et al., 2010). Theseroduce a challenging environment for establishment and growthf vegetation.

Quality green composts can provide nutrients for plants andan also improve soil physical conditions for plant growth (Cogger,005). Such composts contain a wide range of micronutrients,

ncluding magnesium, copper and iron as well as important con-entrations of nitrogen, phosphorus and potassium. The effects ofompost on plant available water (PAW) are dependent on soilype and the properties of the compost such as density and organic

atter content (Curtis and Claassen, 2009). Composts improve theechanical properties of amended soils, including shear strength,

ven at quite large application rates of up to 30% by volume (Hejdukt al., 2012; Puppala et al., 2007). There is conflicting evidencerom Hemmat et al. (2010), who found that at application ratesf 100 t ha−1 cohesion (binding between particles) decreased andnternal friction (frictional resistance between particles) increasedt drier water contents (17.1%), but no differences were found at

wetter water content tested (20.9%) for a silty clay loam soil. Inhe dry condition shear strengths dropped by 60% due to compostmendment, but as most failures occur at small effective stresseswhen capillary cohesion by pore water is small), the similarityf the wet measurements suggests composts alone will not affectlope stability due to decreased shear strength. If the addition ofompost does not affect the shear strength at typical water condi-ions where failures occur, improved root growth associated withhe compost amendment could enhance slope stability.

Increased root length density, RLD, following soil amendmentith organic compounds has been observed in many studies (Hati

t al., 2006; Mosaddeghi et al., 2009). Organic matter and rootength density improve soil aggregation (Materechera et al., 1992),

hich is accentuated by compost amendments (Daynes et al.,013). For single species in laboratory conditions, RLD is related to

ncreased shear strength (Normaniza et al., 2008), although plantuccession (Osman and Barakbah, 2011), root age and environmen-al factors (Loades et al., 2013) are also important.

Using a purpose built experimental slope, we investigated thempact of compost amendment on soil fertility, plant establish-

ent and soil reinforcement by plant roots. Two application rates of5 t ha−1 and 300 t ha−1 were either applied to the surface of the soils a blanket or incorporated into the top 10 cm of the soil surface.he soil was a mineral fill formed from crushed building waste andiscarded soil from demolition projects. We focussed on the ini-ial stage of revegetation within thirteen weeks of planting. Duringhis period, exposed bare soil is particularly vulnerable to shallownstabilities. We hypothesised that improved plant establishment,aused by increased soil fertility from compost, would drive soiltabilisation by plant roots and that this would more than offsetny weakness caused by potentially lower shear strength from theompost addition. The research has direct relevance to improvinghe effectiveness of slope ecoengineering projects.

. Materials and methods

.1. Slope design

An experimental slope was constructed adjacent to Dundee Cityouncil’s composting facility, located at the Riverside Landfill site

moaH

ering 71 (2014) 458–465 459

56◦27′19′′N; 3◦2′32′′W) facing southeast. Material from demoli-ion and construction sites is processed into materials for reuse inhe construction industry at an adjacent site.

The slope required the movement of 1200 t of material that wasuilt up to defined dimensions using heavy earth moving machin-ry (Fig. 1A). The slope was about 40 m wide and had a slope lengthf about 15 m with a 3 m buffer before a road at the toe of the slope.ubble was placed at the base to provide slope stability, and thenoil was layered above to provide a 20◦ slope angle. This is shal-ower than some engineered embankments but ensured safety. Aop layer of soil, produced from the washings of mineral fill waste,as then placed on the surface to a depth of at least 60 cm. Thisaterial was a silty-sand textured recycled soil, consistent with

he surface cover over the remainder of the landfill site. In otherarts of the landfill, surface vegetation had not been seeded buthere was evidence of wind-blown seed resulting in patchy plantrowth.

The surface of the constructed slope had compost applied inreatment blocks (Fig. 1B). We used a randomised block designith 3 replicates of each of the 5 treatments used. Compost was

pplied at 2 rates and either incorporated to a depth of 10 cm byarrowing or left as a surface mulch. The compost used was made

rom plant residues only that were processed using the windrowethod, by Dundee City Council, to produce quality green compost

n accordance with BSI PAS 100 standard before grading to <20 mm.ach tonne, fresh weight, of this compost contained approximately

kg of nitrogen, all organic, 3 kg of phosphorus, as P2O5, 3 kg ofotassium, as K2O and had a dry matter content of 62%. The treat-ents were applied on 2 April 2009 using smaller scale equipment

ommonly used for high-value horticulture. Amenity No. 2 grasseed mix (R M Welch and Sons Ltd, Broughty Ferry, United King-om) was then sown to establish a vegetation cover. Seed mixontained Creeping Red Fescue (Festuca rubra) (50%), Perennialyegrass (Lolium perenne) (30%), Chewing Fescue (F. rubra subsp.ommutata) (10%) and Browntop Bent (Agrostis capillaris) (10%).ollowing seeding, we sampled for 13 weeks between 20th Aprilnd 17th July 2009.

.2. Field sampling and measurements

Monitoring the slope focused on the critical period of vegetationstablishment after slope construction when the surface was bare.t this time compost is likely to be of most benefit in reducingurface erosion and shallow failures by accelerating establishmentf plant cover. Measurements of soil mechanical behaviour, fertilitynd plant growth were taken at pre-determined times between 0nd 12 weeks after completion of slope construction. Rainfall dataere collected at the Mylnefield Weather Station, located less than

km away.On several sampling dates, two sets of cores and a bulk soil sam-

le were taken from the surface of the crest, middle and toe ofach plot. Cores 100 mm × 100 mm were taken perpendicular to thelope for mechanical tests and the abundance of plant roots mea-ured at weeks 1, 7 and 12. Cores 40 mm height × 55 mm diameterere taken for measurement of water retention in the laboratory

t weeks 1, 2, 4, 8, and 12. Bulk soil samples for chemical analysisere taken by combining ten samples collected at random loca-

ions with a 10 cm length × 2.5 cm diameter corer at weeks 0, 1,, 4, 8 and 12. The bulk samples from each plot were combined,assed through a 4 mm sieve and mixed thoroughly.

On each sampling date the percentage cover of vegetation was

easured on a 0.5 m × 0.5 m grid at the crest, middle and toe

f the slope using a square quadrat. Digital images were takennd imported to ImageJ analysis software (National Institutes ofealth, Bethesda, Maryland, USA) where they were converted to

460 S. Donn et al. / Ecological Engineering 71 (2014) 458–465

Fig. 1. (A) Cross-section of the slope showing the layer of rubble that was placed between the pre-existing slope and new fill to provide both drainage and keying to avoidp nt layc ated c1

Rpepmmes

2

aweoNdgP(

wpthp(u

−(pS

2

wm1bsihc4h

5ptrdL

otential slope failure. The surface material is amended with compost; (B) treatmeompost at 35 t ha−1; (3) surface applied compost at 300 ha−1; (4) surface incorpor0 cm depth and 300 t ha−1.

GB stacks. The green layer was converted to a binary image andarticles analysed for percentage cover. Normalised difference veg-tation index, NDVI, was also measured in the field at two locationser plot using a GreenSeekerTM hand held optical sensor unit,odel 505 (NTech Industries Inc., Ukiah, CA). Sward height waseasured on the longest leaf of five plants chosen at random in

ach plot on selected sampling dates. Further details of these mea-urements can be found in McKenzie et al. (2009).

.3. Soil analysis

Nitrate-N, ammonium-N and dissolved organic carbon werenalysed after extracting sub-samples from the bulk soil samplesith 1 M KCl; soil: extract ratio 1:4 on a roller bed for 1 h (Wheatley

t al., 1989). Concentrations were determined colourimetricallyn a segmented flow auto analyser (Skalar Analytical, Breda, Theetherlands). Ten grams of soil were dried overnight at 105 ◦C toetermine the soil water content. Soil-pH was determined with alass electrode in a solution of 10 g of soil in 40 ml of 0.01 M CaCl2.hosphorus, magnesium and potassium were measured in week 0Yara Analytical Services).

Plant available water (PAW), measured in units of volume ofater per volume soil (m3 m−3), is the soil water accessible tolant roots. PAW was measured as the difference in water con-ent between water potentials of −1500 kPa, where any water is

eld in pores too small for roots to access it (permanent wiltingoint), and −5 kPa, the maximum amount of water the soil can holdfield capacity) (Hallett and Bengough, 2013). PAW was measuredsing the 40 mm height × 55 mm diameter cores. Water content at

tb

i

out on the slope. The numbers correspond to (1) no compost; (2) surface appliedompost to 10 cm depth and at 35 t ha−1; and (5) surface incorporated compost to

5 kPa was measured following equilibration on a tension tableELE International, Hemel Hempstead, UK) and at −1500 kPa usingressure plate apparatus (Soil Moisture Equipment Corporation,anta Barbara, USA).

.4. Mechanical behaviour

Cores equilibrated to −5 kPa on a tension table for at least 48 h,ere extracted into a direct-shear rig and sheared at a displace-ent rate of 1 mm min−1 using a Wykenham Farrance (Tring, UK)

00 mm diameter circular direct shear rig. Cores were stored at 4 ◦Cefore equilibration and sheared within 7 days of collection. Cohe-ion and angle of internal friction were assessed on samples takenn week 0 of testing. This involved a 100 mm diameter × 20 mmeight sample, with the shear plane at the centre of the sampleorresponding to 5 cm depth in the field. Normal stresses of 0, 1.25,.3, 9.7 and 20.7 kPa were applied to the surface of the soil usinganging weights.

For other tests the rig was adapted so that the shear depth was cm depth and 5 cm of soil was present above and below the shearlane to minimise the pull-out of roots. Plant stems were trimmedo the surface before testing but left otherwise intact so that rooteinforcement was not affected. The outputs of a 500 N load cell andisplacement transducer (LVDT) were logged continuously usingabView (National Instruments, Austin, Texas, USA) and converted

o load and distance, respectively, from calibrations carried outefore each test event.

The shear stress was calculated from the applied load by divid-ng by the surface area of the shear surface. Peak shear stress was

ngineering 71 (2014) 458–465 461

tmiescctssN

2

Gsct

3

3

aaosilsatmaetc1bpct

Pyt3t

Week

121086420

NO

3 (µg

N g

-1 s

oil)

0

20

40

60

80

100

Fig. 2. Nitrate, NO3, contents measured in soil extracts over the duration of thestudy. The compost amendment treatments are 0 t ha−1 (©), 35 t ha−1 incorporatedto 10 cm depth (�), 35 t ha−1 applied to the surface (�), 300 t ha−1 incorporated to10 cm depth (�) and 300 t ha−1 applied to the surface (�). The standard error isshown.

Wee k

121086420

PA

W (m

3 m-3

)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Fig. 3. Plant available water, PAW, measured as the difference between the watercontents at −5 kPa and −150 kPa. The compost amendment treatments are 0 t ha−1

(©), 35 t ha−1 incorporated to 10 cm depth (�), 35 t ha−1 applied to the surface (�),300 t ha−1 incorporated to 10 cm depth (�) and 300 t ha−1 applied to the surface(

3

TS

S. Donn et al. / Ecological E

aken as the maximum recorded shear stress up to 15 mm displace-ent. No apparent peak was observed for many specimens, which

s common for shear tests on soils containing fibrous roots (Loadest al., 2010). Its value at week 0 was subtracted from peak sheartress at subsequent dates following construction to determine theohesion contributed by plant roots, cr at known times after slopeonstruction. After shear testing the cores were frozen. A 2 cm sec-ion from beneath the shear plane was cut using a diamond saw. Theection was then thawed and washed to measure root length den-ity (RLD) and root diameter using WinRHIZO (Regent Instruments,epean, Canada).

.5. Statistical analysis

Data were checked for normal distribution and analysed usingenStat 15th Edition (VSN International, UK). Regression analy-is within groups evaluated the change in PAW over time andompared differences between treatments. Differences betweenreatments were tested using analysis of variance, ANOVA.

. Results

.1. Soil conditions for plant growth

Key soil chemical properties are listed in Table 1. Compostddition to the constructed slope had a significant impact onll measured properties, generally increasing the concentrationsf available nutrients, total carbon and DOC. Although pH wastatistically different between compost amendment rates andncorporation methods, the range was less than 0.1. Between samp-ing events on weeks 0, 1, 2, 4, 8 and 12 there were statisticallyignificant differences in K (P < 0.05) and C (P < 0.01) with no inter-ction with treatment, whereas the concentrations varied overime for Mg (P < 0.001) with a significant interaction with treat-

ent (P < 0.05). Nutrient levels increased with addition of compostnd this effect was greater than the variation seen in nutrient lev-ls over time within treatments. P and DOC did not vary withime. NO3–N concentrations decreased rapidly over time, to theoncentrations of no compost amendment for all treatments after2 weeks (P < 0.001; Fig. 2). There was a significant interactionetween treatment and time for NO3 (P < 0.001). With all com-ounds measured, the concentrations were greater for 300 t ha−1

ompost compared to 35 t ha−1, and greater for surface applicationhan for incorporation.

At week 1, only the 300 t ha−1 surface treatment had a greaterAW than the 0 t ha−1 treatment (P < 0.05; Fig. 3). Regression anal-

sis tested changes over time and differences in trends betweenreatments. Whilst PAW remained unchanged for 0 t ha−1 and5 t ha−1 incorporated compost, it increased over time for all otherreatments (P < 0.01, R2 = 0.44).

3ftc

able 1oil chemical properties at the start of the experiment in week 0.

Compost amendment treatment

0 t ha−1 35 t ha−1

Surface

Carbon [g 100 g−1] 2.35 ± 0.17 2.55 ± 0.11

Nitrogen [g 100 g−1] 0.243 ± 0.043 0.338 ± 0.029

Phosphorus [�g g−1] 27.78 ± 0.84 33.17 ± 0.88

Potassium [�g g−1] 250.8 ± 16.8 330.7 ± 14.7

Magnesium [�g g−1] 284.9 ± 6.8 299.5 ± 3.6

pH 7.98 ± 0.57 7.95 ± 0.26

Dissolved organic carbon [�g g−1] 39.01 ± 3.57 73.71 ± 4.62

�).The standard error and regression lines are shown.

Soil water content was much greater in plots amended with00 t ha−1 compost and marginally greater in plots amended with

5 t ha−1 in comparison to 0 t ha−1 (P < 0.001; Fig. 4), for both sur-ace applied and incorporated treatments. Water status varied overime (P < 0.001), with soils remaining relatively dry until the pre-ipitation event that occurred in week 4.

P

35 t ha−1 300 t ha−1 300 t ha−1

Incorporated Surface Incorporated

2.10 ± 0.08 5.92 ± 0.11 4.39 ± 0.26 <0.0010.248 ± 0.044 0.569 ± 0.044 0.390 ± 0.037 <0.00131.28 ± 0.67 75.56 ± 3.25 54.33 ± 1.87 <0.001293.2 ± 11.4 1208 ± 69.0 724.2 ± 35.14 <0.001295.1 ± 5.5 440.5 ± 14.4 362.9 ± 8.3 <0.0017.96 ± 0.03 7.90 ± 0.01 7.89 ± 0.04 0.00152.94 ± 4.19 268.7 ± 20.9 140.4 ± 6.66 <0.001

462 S. Donn et al. / Ecological Engineering 71 (2014) 458–465

Week

121086420

Soi

l Wat

er C

onte

nt (g

100

g-1)

0

10

20

30

40

Pre

cipi

tatio

n (m

m)

0

10

20

30

40

50

Fig. 4. Soil water content and precipitation over the first 12 weeks of vegeta-tion establishment. The compost amendment treatments are 0 t ha−1 (©), 35 t ha−1

i −1 −1

is

3

(ttwpci0rtNi3ttmd

Fsl

Wee k

121086420

ND

VI

0.0

0.2

0.4

0.6

0.8

1.0

Fig. 6. NDVI during plant establishment measured using a Greenseeker. Measure-ments were taken at two positions per plot and the mean of three treatmentreplicates is shown. The compost amendment treatments are 0 t ha−1 (©), 35 t ha−1

iis

aht

twtTspa

3

u

ncorporated to 10 cm depth (�), 35 t ha applied to the surface (�), 300 t hancorporated to 10 cm depth (�) and 300 t ha−1 applied to the surface (�). Thetandard error is shown.

.2. Vegetation

Vegetation establishment was initially slow in all treatmentsFig. 5), possibly due to delayed germination in the dry conditions athe start of the experiment, with only 32 mm rainfall in total duringhe first three weeks post seeding (Fig. 4). At week 4, surface coveras significantly greater in the 300 t ha−1 plots and the 35 t ha−1

lots, in both surface applied and incorporated treatments, than theontrol (P < 0.001). By weeks 8 and 12, vegetation cover decreasedn the order 300 t ha−1 greater than the 35 t ha−1 greater than

t ha−1 (P < 0.001). At both application rates, surface applicationesulted in greater surface cover by vegetation than incorpora-ion of compost. In addition to greater surface vegetation cover,DVI was greater with 300 t ha−1 applications indicating health-

er, less-stressed plant quality or “greenness” (P < 0.001; Fig. 6). At5 t ha−1 the surface application plots show greater “greenness”han incorporated. Improved plant establishment probably drove

his difference. Sward height was greater in the 300 t ha−1 treat-

ents than in the 35 t ha−1 treatments (P < 0.001), which were notifferent from the control (Fig. 7), indicating better plant growth

Compost Amendment (t ha-1)

300 Surf.300 Inc.35 Surf.35 Inc.0

Veg

etat

ion

Cov

er (m

2 m-2

)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

ig. 5. Vegetation cover in week 4 (black), 8 (light grey), and 12 (dark grey). Thetandard error is shown. (For interpretation of the references to color in this figureegend, the reader is referred to the web version of this article.)

mbw

Fp(a

ncorporated to 10 cm depth (�), 35 t ha−1 applied to the surface (�), 300 t ha−1

ncorporated to 10 cm depth (�) and 300 t ha−1 applied to the surface (�). Thetandard error is shown.

s well as establishment in 300 t ha−1 compost treatments. Swardeights in the 300 t ha−1 surface application were greater thanhose from incorporated plots.

At week 7, RLD was greater in 300 t ha−1 plots than the con-rol, with 35 t ha−1 treatments intermediate (P = 0.002; Fig. 8). Byeek 12, RLD was greater in 35 t ha−1 treatments than in the con-

rol, with the 300 t ha−1 treatments remaining significantly longer.here was no significant difference between the incorporated andurface applied plots on root growth on any sampling dates. Com-ost application did not affect root diameter (P = 0.09), with theverage diameter in week 12 of 0.48 mm ± 0.05 mm.

.3. Mechanical properties

Direct shear tests of intact soil cores collected at week 0 andnder varying normal stresses demonstrated that compost amend-

ent did not influence either cohesion or internal angle of friction,

ased on Mohr–Coulomb failure criterion (P > 0.05). As the shearas at 5 cm beneath the soil surface and compost was incorporated

Wee k

121086420

Sw

ard

heig

ht (m

m)

0

20

40

60

80

100

120

140

160

ig. 7. Sward height over the first 12 weeks of vegetation establishment. The com-ost amendment treatments are 0 t ha−1 (©), 35 t ha−1 incorporated to 10 cm depth�), 35 t ha−1 applied to the surface (�), 300 t ha−1 incorporated to 10 cm depth (�)nd 300 t ha−1 applied to the surface (�). The standard error is shown.

S. Donn et al. / Ecological Engineering 71 (2014) 458–465 463

Compost Amendment (t ha-1)

300 Surf.300 Inc.35 Surf.35 Inc.0

Roo

t len

gth

dens

ity (1

04 m m

-3)

0

10

20

30

40

50

60

Fig. 8. Root length density, RLD, in week 1 (black), 7 (light grey), and 12 (dark grey)measured on a 2 cm slice of soil taken beneath the failure surface from the shear testcores. The standard error is shown. (For interpretation of the references to color inthis figure legend, the reader is referred to the web version of this article.)

Table 2Root cohesion, cr measured at 5 cm depth in the soil cores for incorporated composttreatments at different times from the start of the experiment were used in thesimulation. Samples are limited to incorporated compost treatments.

Compost amendment treatment

0 t ha−1 (kPa) 35 t ha−1 (kPa) 300 t ha−1 (kPa)

Week 1 0 0 0

tdica

cf(fsmr3

4

nortsiab

sgah

Compost Amend ment (t ha-1)

300 Surf.300 Inc.35 Surf.35 Inc.0

Pea

k sh

ear s

tress

(kP

a)

0

2

4

6

8

10

12

14

16

Fig. 9. Peak shear stress of extracted soil cores in week 1 (black), 7 (light grey), and12 (dark grey). The standard error is shown. Cores were 100 mm diameter and takentt

ifvpeamamtatlw1andwau

pnmrdaglcao

ifaI

Week 7 1.2 3.1 7.9Week 12 2.1 2.3 8.1

o 10 cm depth, compost would have been present at the shearepth. We assume that the samples were fully drained during test-

ng and at low applied principal stresses they failed with minimalhange to water potential. This results in cohesion, c′ = 1.9 kPa andngle of friction, ϕ = 37.8◦.

At week 1 peak shear stress of the treatments with incorporatedompost at either 35 t ha−1 or 300 t ha−1 were still not differentrom the 0 t ha−1 (Table 2). Peak shear stress increased with timeP < 0.001) as vegetation became established, suggesting root rein-orcement of the soil (Fig. 9). In week 7, 300 t ha−1 treatmentshowed significantly greater peak stress, with 35 t ha−1 treat-ents stronger than the control. In week 12, 300 t ha−1 treatments

emained significantly stronger than all other treatments though5 t ha−1 treatments were not different to the controls (P < 0.001).

. Discussion

Quality green compost applied at the rates used here hado direct impact on Mohr–Coulomb shear properties (c and ϕ)f the soil investigated but improved plant growth and thusoot reinforcement. At the largest application rate, 300 t ha−1,he effects were greatest, with surface applied compost treatedlopes performing marginally better than plots with compostsncorporated in the top 10 cm. For engineering projects, surfacepplication is easier and more cost-effective to apply, so likely toe the method of choice.

A short-term benefit of compost application to constructed

lopes is the provision of nutrients that promote rapid plantrowth. At a different location, Persyn et al. (2007) found that

5 cm layer of compost, which is similar to the 300 t ha−1usedere, performed as well as a similar depth of imported topsoil at

bnto

o 100 mm depth. (For interpretation of the references to color in this figure legend,he reader is referred to the web version of this article.)

mproving vegetation establishment. The N, P, K and Mg suppliedrom the compost will be a major cause of the 10–20 fold increase inegetation cover seen over the 12 weeks, in the 300 t ha−1 amendedlots compared to the controls. The measured increases in nutri-nts were significant for 35 t ha−1 added compost (Table 1), andlso increased vegetation cover and other measures of plant perfor-ance over 12 weeks. The nutrient concentrations at 35 t ha−1 of

pplied compost are still comparable, in gross terms, to those thatight be used agronomically. The total amounts of N, P and K in

he 35 t ha−1 application are approximately 290 kg of N, 76 kg of Pnd 106 kg K per hectare. However, applied amounts of N are abovehose allowable under the new Nitrogen Vulnerable Zone guide-ines (Scottish Government, 2013), if applied yearly, but allowable

ithin the 2 year rolling permissible limits of 500 kg ha−1. About0% of the N and 50 to 60% of the P and K might be in readilyvailable forms, but the longevity of these nutrient amendmentseeds greater investigation. NO3–N concentrations in the soilropped rapidly during the 4 week period after application andithin 12 weeks concentrations were similar between amended

nd unamended plots. Much of this NO3–N will have been takenp by the plants, but there are potential environmental losses.

Despite being formed from recycled materials, the unamendedlots had appreciable concentrations of N that were similar to aearby agricultural field sampled in the winter approximately 3onths after the slope was constructed (Sun et al., 2011). Soils

eclaimed in mining projects can have N concentrations belowetectable levels (Leifeld et al., 2001), so the relative impact ofny compost addition on vegetation establishment could be muchreater than for soils with already high nutrient status. Neverthe-ess, after 12 weeks growth there was still <5% of the surface areaovered by vegetation in the unamended plots, in comparison tobout 60% for 300 t ha−1 compost, reflecting the relative abundancef required nutrients for grass growth.

Amendment with compost had the expected impact of increas-ng soil carbon concentrations to levels greater than in surroundingarmland (Sun et al., 2011). Appreciable increases in DOC werelso measured, which is easily mineralised by soil microbes.n soil engineering or restoration projects, microbial and other

iological activity is often poor but important to the cycling ofutrients for plant growth (Ohsowski et al., 2012). It was beyondhe scope of this study to measure soil microbial populations, butther studies have shown increased microbial biomass, activity

4 ngine

(iscetsn

iCmdCwao3igw

wAFrwcaatwostIw£ss

bObgflr

5

rrsttrfiwiArr

os

os

scn

s

r

em

u

f

ap

p

abvmr

A

TEt

R

B

B

C

C

C

D

d

D

E

F

64 S. Donn et al. / Ecological E

Ros et al., 2003) and functional diversity (Gomez et al., 2006)n response to compost addition. Furthermore, on agriculturaloils significant shifts in nitrifier populations were observed withompost amendments (Griffiths et al., 2010). There is increasingvidence that root exudates produced by vegetation will enhancehe mineralisation of recalcitrant carbon, including from compostources (Paterson et al., 2008). This will influence the release ofutrients and the longer-term development of soil structure.

Improved soil structure over time accounts for the increasesn PAW we observed between the start and end of monitoring.urtis and Claassen (2009) found that following compost amend-ent, PAW either increased or decreased dependent on soil type. A

ecrease in PAW contradicts many previous studies, but Curtis andlaassen (2009) argued that other authors measured gravimetricater content, which is biased by the smaller density of compost

mended soils. Shortly after compost was added to our plots webserved no differences in PAW apart from a small increase for00 t ha−1 compost applied to the surface. Over time, however,

mproved PAW likely demonstrates the combined processes of rootrowth, compost mineralisation and soil aggregation on improvedater retention (Hallett et al., 2009; Zhang et al., 2005).

A direct consequence of improved vegetation establishmentas the mechanical reinforcement of the soil by plant roots.dding 300 t ha−1 compost almost quadrupled root cohesion.or soils at low effective stress near to saturation, this shouldesult in increased shallow slope stability at this critical conditionhen failures often occur. As a result, steeper slopes could be

onstructed through marrying compost to improve vegetationnd direct reinforcement by plant roots. It is possible to maken economic assessment of the value of compost versus topsoilo improve the fertility of constructed slopes. A 1 km long, 10 mide slope is 1 ha in area. Based on the current commercial cost

f compost at £10 t−1, transport to a site of £10 t−1 km−1, and apreading cost of £100 km−1 for a 10 m slope, adding 35 t ha−1

o 300 t ha−1 compost would add £800 to £6100 per kilometre.n comparison, 5 cm of topsoil with a density of 1300 kg m−3

eighs 650 t, so transporting and applying it would cost at least13,100 km−1, based on a topsoil cost of £10 t−1 and similarhipping and application costs for compost. The additional cost forpreading and preparation may increase the difference further.

The positive benefits of compost on geotechnical slopes haveeen reported for different locations and by other researchers (dena and Osorio, 2006; Osorio and de Ona, 2006; Persyn et al., 2004),ut this is the first direct evidence of compost promoting plantrowth and consequently improved root cohesion of shallow sur-ace soil. Further research is needed to explore performance overonger time periods and of woody vegetation with much deeperoot systems than grasses.

. Conclusions

The use of quality green compost in slope construction willesult in significant reinforcement of these soil slopes at depthseached by plant roots. This occurs for several reasons. Firstly,oil fertility is improved, with increased concentrations of essen-ial plant nutrients including N, P, K and Mg. Most significant ishe improvement in vegetation establishment and hence the plantoot development which reinforces the slope. The greatest bene-ts to root reinforcement were observed after surface applicationith the relatively large inputs of 300 t ha−1 of compost, although

ncorporation at this rate also significantly increased slope stability.pplications of compost at the lower level, 35 t ha−1, also improvedoot reinforcement but not to such a great extent. No adverse envi-onmental effects were seen in any of the treatments, with data

F

G

ering 71 (2014) 458–465

n erosion measurements and soil hydrology to be presented in aubsequent publication.

GlossaryAngle of Friction, ϕ: one parameter in Mohr–Coulomb analysis

f the shear strength of soil. It is influenced by the roughness andize distribution of particles.

Cohesion, c: one parameter in Mohr–Coulomb analysis of thehear strength of soil. It is based on the bonding between soil parti-les determined from the relationship between shear strength andormal stress.

DOC–C: Dissolved organic carbon, the carbon dissolved in theoil water expressed as mass per unit mass of soil.

Mohr–Coulomb: a mathematical model describing theesponse of soil to shear stress as well as normal stress.

NDVI: Normalised difference vegetation index a measure of veg-tation colour or health derived from the colour of the materialeasured as reflectance of 2 known wavelengths.Nitrate–N: the amount of nitrogen as nitrate in the sample,

sually expressed as mass per unit mass of soil.Normal stress, �: vertical confining pressure applied to the sur-

ace of soil.Plant available water, PAW: the volumetric water content

ccessible to plants between field capacity (−5 kPa) and wiltingoint (−150 kPa).

Root cohesion, cr: additional shear strength contributed bylant roots.

Slip4Ex: a computer programme for routine stability analysisnd the assessment of the contribution of vegetation to slope sta-ility. Soil reinforcement by geosynthetic layers or anchors, andegetation effects of enhanced cohesion, changed water pressures,ass of vegetation, wind forces and root reinforcement forces are

eadily included in the analysis.

cknowledgements

This study was funded by WRAP (Project code: OBF009-047).he James Hutton Institute receives funding from the Rural andnvironment Science and Analytical Services Division (RESAS) ofhe Scottish Government.

eferences

azzoffi, P., Pellegrini, S., Rocchini, A., Morandi, M., Grasselli, O., 1998. The effect ofurban refuse compost and different tractors tyres on soil physical properties,soil erosion and maize yield. Soil Tillage Res. 48, 275–286.

oyer, S., Wratten, S.D., 2010. The potential of earthworms to restore ecosystemservices after opencast mining—a review. Basic Appl. Ecol. 11, 196–203.

ao, S.X., Xu, C.L., Ye, H.H., Zhan, Y., Gong, C., 2010. The use of air bricks for plantingroadside vegetation: a new technique to improve landscaping of steep roadsidesin China’s Hubei Province. Ecol. Eng. 36, 697–702.

ogger, C.G., 2005. Potential compost benefits for restoration of soils disturbed byurban development. Compost Sci. Util. 13, 243–251.

urtis, M.J., Claassen, V.P., 2009. Regenerating topsoil functionality in four drasticallydisturbed soil types by compost incorporation. Restor. Ecol. 17, 24–32.

aynes, C.N., Field, D.J., Saleeba, J.A., Cole, M.A., McGee, P.A., 2013. Development andstabilisation of soil structure via interactions between organic matter, arbuscu-lar mycorrhizal fungi and plant roots. Soil Biol. Biochem. 57, 683–694.

e Ona, J., Osorio, F., 2006. Application of sludge from urban wastewater treatmentplants in road’s embankments. J. Hazard. Mater. 131, 37–45.

iacono, M., Montemurro, F., 2010. Long-term effects of organic amendments onsoil fertility. A review. Agron. Sustainable Dev. 30, 401–422.

dwards, L., Burney, J.R., Richter, G., MacRae, A.H., 2000. Evaluation of compost andstraw mulching on soil-loss characteristics in erosion plots of potatoes in PrinceEdward Island, Canada. Agric. Ecosyst. Environ. 81, 217–222.

aucette, L.B., Scholl, B., Beighley, R.E., Governo, J., 2009. Large-scale performanceand design for construction activity erosion control best management practices.J. Environ. Qual. 38, 1248–1254.

ernandez-Getino, A.P., Perez, J., Albarran, M.D., Lammerding, D.M., Walter, I., 2012.Restoration of abandoned, degraded agricultural soil using composted biosolid:influence on selected soil properties. Arid Land Res. Manage. 26, 200–210.

omez, E., Ferreras, L., Toresani, S., 2006. Soil bacterial functional diversity as influ-enced by organic amendment application. Bioresour. Technol. 97, 1484–1489.

ngine

G

G

H

H

H

H

H

H

L

L

L

M

M

M

M

M

M

M

N

O

O

O

P

P

P

P

R

S

S

S

V

W

S. Donn et al. / Ecological E

riffiths, B.S., Ball, B.C., Daniell, T.J., Hallett, P.D., Neilson, R., Wheatley, R.E., Osler,G., Bohanec, M., 2010. Integrating soil quality changes to arable agriculturalsystems following organic matter addition, or adoption of a ley-arable rotation.Appl. Soil Ecol. 46, 43–53.

ruenewald, H., Brandt, B.K.V., Schneider, B.U., Bens, O., Kendzia, G., Huttl, R.F.,2007. Agroforestry systems for the production of woody biomass for energytransformation purposes. Ecol. Eng. 29, 319–328.

allett, P.D., Feeney, D.S., Bengough, A.G., Rillig, M.C., Scrimgeour, C.M., Young, I.M.,2009. Disentangling the impact of AM fungi versus roots on soil structure andwater transport. Plant Soil 314, 183–196.

allett, P.D., Bengough, A.G., 2013. Managing the soil physical environment forplants. In: Gregory, P.J., Nortcliff, S. (Eds.), Soil Conditions and Plant Growth.Wiley-Blackwell, London/Oxford, UK, pp. 238–268.

ansen, N.E., Vietor, D.M., Munster, C.L., White, R.H., Provin, T.L., 2009. Runoff waterquality from constructed roadside soils receiving large volume-based compostapplications. Trans. ASABE 52, 793–800.

ati, K.M., Mandal, K.G., Misra, A.K., Ghosh, P.K., Bandyopadhyay, K.K., 2006. Effectof inorganic fertilizer and farmyard manure on soil physical properties, rootdistribution, and water-use efficiency of soybean in Vertisols of central India.Bioresour. Technol. 97, 2182–2188.

ejduk, S., Baker, S.W., Spring, C.A., 2012. Evaluation of the effects of incorporationrate and depth of water-retentive amendment materials in sports turf construc-tions. Acta Agric. Scand. Sect. B: Soil Plant Sci. 62, 155–164.

emmat, A., Aghilinategh, N., Sadeghi, M., 2010. Shear strength of repackedremoulded samples of a calcareous soil as affected by long-term incorporationof three organic manures in central Iran. Biosyst. Eng. 107, 251–261.

eifeld, J., Siebert, S., Kogel-Knabner, I., 2001. Stabilization of composted organicmatter after application to a humus-free sandy mining soil. J. Environ. Qual. 30,602–607.

oades, K.W., Bengough, A.G., Bransby, M.F., Hallett, P.D., 2010. Planting densityinfluence on fibrous root reinforcement of soils. Ecol. Eng. 36, 276–284.

oades, K.W., Bengough, A.G., Bransby, M.F., Hallett, P.D., 2013. Reinforcement ofsoil by fibrous roots. In: Timlin, D., Ahuja, L. (Eds.), Enhancing Understandingand Quantification of Soil-Root Growth Interactions. ASA-SSSA-CSSA, Madison,WI, pp. 276–284.

allik, A.U., Karim, M.N., 2008. Roadside revegetation with native plants: exper-imental seeding and transplanting of stem cuttings. Appl. Veg. Sci. 11,547–554.

arando, G., Jimenez, P., Hereter, A., Julia, M., Ginovart, M., Bonmati, M., 2011. Effectsof thermally dried and composted sewage sludges on the fertility of residual soilsfrom limestone quarries. Appl. Soil Ecol. 49, 234–241.

aterechera, S.A., Dexter, A.R., Alston, A.M., 1992. Formation of aggregates by plant-roots in homogenized soils. Plant Soil 142, 69–79.

cKenzie, B.M., Bengough, A.G., Hallett, P.D., Thomas, W.T.B., Forster, B., McNicol,J.W., 2009. Deep rooting and drought screening of cereal crops: a novel field-based method and its application. Field Crops Res. 112, 165–171.

ickovski, S.B., Buss, K., McKenzie, B.M., Sokmener, B., 2013. Laboratory study on

the potential use of recycled inert construction waste material in the substratemix for extensive green roofs. Ecol. Eng. 61, 706–714.

ontoro, J.A., Rogel, J.A., Querejeta, J., Diaz, E., Castillo, V., 2000. Three hydro-seedingrevegetation techniques for soil erosion control on anthropic steep slopes. LandDegrad. Dev. 11, 315–325.

W

Z

ering 71 (2014) 458–465 465

osaddeghi, M.R., Mahboubi, A.A., Safadoust, A., 2009. Short-term effects of tillageand manure on some soil physical properties and maize root growth in a sandyloam soil in western Iran. Soil Tillage Res. 104, 173–179.

ormaniza, O., Faisal, H.A., Barakbah, S.S., 2008. Engineering properties ofLeucaena leucocephala for prevention of slope failure. Ecol. Eng. 32,215–221.

hsowski, B.M., Klironomos, J.N., Dunfield, K.E., Hart, M.M., 2012. The potential ofsoil amendments for restoring severely disturbed grasslands. Appl. Soil Ecol. 60,77–83.

sorio, F., de Ona, J., 2006. Using compost from urban solid waste to prevent ero-sion in road embankments. J. Environ. Sci. Health A: Environ. Sci. Eng. 41,2311–2327.

sman, N., Barakbah, S.S., 2011. The effect of plant succession on slope stability.Ecol. Eng. 37, 139–147.

aterson, E., Thornton, B., Midwood, A.J., Osborne, S.M., Sim, A., Millard, P., 2008.Atmospheric CO2 enrichment and nutrient additions to planted soil increasemineralisation of soil organic matter, but do not alter microbial utilisation ofplant- and soil C-sources. Soil Biol. Biochem. 40, 2434–2440.

ersyn, R.A., Glanville, T.D., Richard, T.L., Laflen, J.M., Dixon, P.M., 2004.Environmental effects of applying composted organics to new high-way embankments: Part 1. Interrill runoff and erosion. Trans. ASAE 47,463–469.

ersyn, R.A., Richard, T.L., Glanville, T.D., Laflen, J.M., Dixon, P.M., 2007. Evaluationof revegetation from blanket applied composts on a highway construction site.Appl. Eng. Agric. 23, 631–635.

uppala, A.J., Pokala, S.P., Intharasombat, N., Williarnmee, R., 2007. Effectsof organic matter on physical, strength, and volume change propertiesof compost amended expansive clay. J. Geotech. Geoenviron. Eng. 133,1449–1461.

os, M., Hernandez, M.T., Garcıa, C., 2003. Soil microbial activity after restoration ofa semiarid soil by organic amendments. Soil Biol. Biochem. 35, 463–469.

aid-Pullicino, D., Massaccesi, L., Dixon, L., Bol, R., Gigliotti, G., 2010. Organic matterdynamics in a compost-amended anthropogenic landfill capping-soil. Eur. J. SoilSci. 61, 35–47.

cottish Government, 2013. The Action Programme for Nitrate Vulnerable Zones(Scotland) Amendment Regulations. In: Scottish Statutory Instruments No. 123,Environmental Protection: Agriculture & Water. The Stationery Office Limited,London, 5.

un, B., Hallett, P.D., Caul, S., Daniell, T.J., Hopkins, D.W., 2011. Distribution of soilcarbon and microbial biomass in arable soils under different tillage regimes.Plant Soil 338, 17–25.

aidya, G.S., Shrestha, K., Khadge, B.R., Johnson, N.C., Wallander, H., 2008. Organicmatter stimulates bacteria and arbuscular mycorrhizal fungi in Bauhinia pur-purea and Leucaena diversifolia plantations on eroded slopes in Nepal. Restor.Ecol. 16, 79–87.

arman, P.R., Rodd, A.V., Hicklenton, P., 2009. The effect of MSW compost and fer-tilizer on extractable soil elements and the growth of winter squash in Nova

Scotia. Agric. Ecol. Environ. 133, 98–102.

heatley, R., MacDonald, R., Smith, A., 1989. Extraction of nitrogen from soils. Biol.Fert. Soil. 8, 189–190.

hang, B., Horn, R., Hallett, P., 2005. Mechanical resilience of degraded soil amendedwith organic matter. Soil Sci. Soc. Am. J. 69, 864–871.