An ecosystem services approach to the quantification of shallow massmovement erosion and the value of soil conservation practices

E.J. Dominati a,n, A. Mackay a, B. Lynch b,1, N. Heath b,1, I. Millner b,1

a AgResearch, Grasslands Research Centre, Tennent Drive, Private Bag 11008, Palmerston North 4442, New Zealandb Hawkes Bay Regional Council, 159 Dalton Street, Private Bag 6006, Napier 4142, New Zealand

a r t i c l e i n f o

Article history:Received 28 March 2014Received in revised form25 June 2014Accepted 26 June 2014Available online 5 August 2014

Keywords:Soil changeEcosystem servicesNatural capitalErosion recoverySoil conservationBenefit cost analysis

a b s t r a c t

This study characterises the loss of ecosystem services from a grazed pasture following shallow massmovement erosion and subsequent recovery of services. The influence of space-planted trees,a soil conservation practice, on the provision of services, was also assessed. The economic value ofthe services provided by an uneroded steep pasture grazed by sheep and cattle was estimated at NZD3717 ha�1 yr�1. This value dropped by 65% when the topsoil was lost in a single shallow massmovement. Fifty years after erosion, the services only recovered to 61% of uneroded value. In contrast,the same landscape type planted with soil conservation trees provided, after 20 years, additional (þ22%in dollar value) services from the similar unprotected landscape.

A benefit cost analysis of soil conservation practices showed planting conservation trees is onlyprofitable if the trees are harvested for timber (age 20), and low discount rates (o5%) are used. Whenthe economic value of the extra services from conservation trees is included in the BCA, the Net PresentValue of the investment is greatly positive at discount rates ranging from 0% to 10%. Analysis of thisecological infrastructure investment using an ecosystem service approach offers new insights forresource managers and policy makers.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

New Zealand is one of the only places in the world where hillcountry is grazed all year round. Almost 70% of the country hasslopes greater than 12 deg., and is commonly called ‘hill country’.Geologically, the majority of the North Island, where this studytakes place, is developed on soft rock and crushed soft rock terrain(McIvor et al., 2011). These combined with warm sub-tropical tocool temperate climates make the North Island hill country highlyprone to shallow landslides, earthflow and gully erosion (Basheret al., 2008; McIvor et al., 2011). Nowadays, historical deforestationassociated with pastoral land use, also plays a major role indetermining erosion risk.

The Water and Soil Conservation Act was passed in NewZealand in 1941 to address hill country erosion associated withpost-European settlement and deforestation. Catchment Boards,directed by central government policies, were tasked with soil andwater conservation until 1988. In 1988, Catchment Boards wereabsorbed into Regional or Unitary Councils responsible for broader

natural resource management, including soil erosion and floodcontrol under the Resource Management Act (RMA) of 1991. Eachyear hill country erosion is estimated to cost between NZD 100 to150 million (Eastwood et al., 2001). Part of this is through lostpasture production and nutrients (MfE, 2007), but does not includean estimate of the loss of soil natural capital stocks (Dominati et al.,2010). The investment in soil conservation continues today, aserosion remains a challenge threatening the long-term sustain-ability of agro-ecosystems. This is not unique to New Zealand but athreat to food security in many regions of the world (McBratney etal., 2014), heightened by uncertainties surrounding future climates.

Soil conservation practices aim to reduce the risk of soil erosionin hill and steep land country, downstream costs associated withsediment loadings in waterways, and damage to productive farm-land and towns through siltation. In New Zealand, erosion controlmeasures for pastoral hill country are established in the presenceof the grazing animal, with permanent retirement from grazingrecommended only in the most extreme situations. Tree-basedcontrol measures, which stabilise mass flows, are most of the time,the only affordable option on the scale required. Poplar (Populusspp.) and willow (Salix spp.) are two of the most suitable and mostused tree species (McIvor et al., 2011).

Current evaluation of soil conservation policies are largelylimited to the assessment of the reduction in soil erosion, soilloss, sediment, impacts on productive capacity and downstream

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/ecoser

Ecosystem Services

http://dx.doi.org/10.1016/j.ecoser.2014.06.0062212-0416/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author. Tel: þ64 63518216.E-mail addresses: estelle.dominati@agresearch.co.nz (E.J. Dominati),

alec.mackay@agresearch.co.nz (A. Mackay), barry@hbrc.govt.nz (B. Lynch),nathan@hbrc.govt.nz (N. Heath), millner@hbrc.govt.nz (I. Millner).

1 Tel.: þ64 68359209.

Ecosystem Services 9 (2014) 204–215

community. However, evaluation needs to be broader in order tocapture the wider benefits of soil conservation. For example thevalue of the full range of ecosystem services (ES), above- andbelow-ground, lost following shallow mass movements need to beconsidered, beyond the loss of selected provisioning services. Untilthe full range of services is considered in the analysis, the cost ofland degradation and loss of natural capital stocks will remainunknown and the full value of an ecological infrastructure invest-ment in soil conservation practices will not be available for landuse decision-making.

The purpose of this study is to test and evaluate an ecosystemservices approach for the quantification and economic valuationof the multiple benefits, provided by soil conservation practicesusing a study area on the East Coast of the North Island of NewZealand.

In April 2011, Hawke's Bay on the East Coast of the North Islandwas affected by a heavy rain storm (200 to 650 mm of rain in 12hours). Subsequent widespread shallow mass movement erosionoccured on hill slopes, including slips and reactivated earth flows, aswell as gullying, along a 250 km coastal strip predominantly underpermanent pasture grazed by sheep and cattle. Following thatstorm, Hawke's Bay Regional Council used satellite imagery toestimate the proportion of land affected by landslides. Overall43 km2 (4300 ha) of bare ground was classified from a total areaof 5900 km2, including 86% new bare ground resulting from thestorm (Jones et al., 2011). Estimates of damage to infrastructure andland, personal and commercial property was NZD 39 million.

As part of a wider analysis, Hawke's Bay Regional Council wasinterested in investigating the long-term implications of the stormevent on the region's natural resource base. This provided anopportunity to quantify and value the ecosystem services lost as aconsequence of shallow mass movement erosion and at the sametime the benefits of existing soil conservation practices. While thisstudy produces economic values for ecosystem services, the mainpurpose of the research was to explore the merits of an ecosystemservice approach in providing resource management decision makerswith new insight into the long-term consequences of erosion and thecosts and benefits of an ecological infrastructure investment. Theobjective was to determine if an ecosystem services approach offers anew tool for amending existing and shaping future policy.

In order to look at the impacts of shallow mass movementerosion and soil conservation practices, the contribution of soils tothe provision of all ecosystem services was assessed using thetheoretical natural capital-ecosystem service framework ofDominati et al., (2010). That framework builds on the millenniumecosystem assessment (MEA, 2005), and differentiates ecologicalprocesses, particularly soil processes, from the flow of ecosystemservices, and integrates the relationships between natural capitalstocks, the impacts of external drivers, and the provision ofecosystem services with human needs. Within the context of thisstudy, the framework is used to account for below and aboveground contributions of natural capital stocks to the provision ofecosystem services.

The ecosystem services considered include provisioning servicessuch as the provision of food (quantity and quality), wood andfibre, the provision of support for human infrastructures and farmanimals, and the provision of shade and shelter for livestock.The regulating services include flood mitigation, the filtering ofnutrients and contaminants, the decomposition of wastes, netcarbon accumulation in soils and conservation trees, nitrous oxideregulation, methane oxidation and the regulation of pest anddisease populations. Cultural services (the nonmaterial benefitspeople obtain from ecosystems) (MEA, 2005) are recognisedbut not considered in this study. Their non-biophysical nature,requires the use of very different techniques for quantification andvaluation and as such are outside the scope of this study.

2. Methodology

2.1. Study site: East Coast hill country sheep and beef operation

The dominant land use in Hawke's Bay region on the East Coastof the North Island of New Zealand is sheep and beef farming.Average farm monitoring data (MPI, 2012) for a summer-dry hillcountry breeding and semi-finishing sheep and beef operationwere used for this study (Table 1). The farm characteristics includea 70:30 sheep to cattle ratio, 130% lambing rate, stocking rate of 10stock units2 per ha, and pasture growth of 9 t of dry matter/ha/yr.Average yearly Rainfall is 1000 mm, and the climate is described assummer-dry. Soil and landscape information was provided byHawke's Bay Regional Council (Table 1). The farm has twodominant landscape units described as the rolling landscape unit,being flat to easy rolling country, and the steep landscape unit,being moderate to steep hill country.

To answer the study objectives, the following steps were taken:

� Quantify and value the provision of ecosystem services for asheep and beef operation (Section 2.3).

� Quantify the impact of erosion on soil properties, the flow ofecosystem services and their recovery (Section 2.4).

� Quantify and value the impact of soil conservation practices onthe flow of ecosystem services from a sheep and beef farm(Section 2.5).

� Undertake a benefit-cost analysis of an investment in soilconservation (Section 2.7).

For all the steps mentioned above, the provision of ecosystemservices is calculated at the paddock scale (the ecosystem servicessource area) on a 1 ha basis.

To quantify the provision of ecosystem services at the paddockscale (1 ha), information from existing data bases and tools thatsupport existing planning in the region, including data collected aspart of soil quality monitoring, land use capability class maps(Lynn et al., 2009) and the OVERSEERs nutrient budget model

Table 1Farm size, landscape type, soil classification and properties, and nutrient inputs to asheep and beef operation.

Rolling landscape unit Steep landscape unit

Area (ha) 255 (45%) 315 (55%)Slope class Easy hill Steep hillSlope 16–251 4261Relative productivity 1.6 1NZ soil classification

Order Brown PallicGroup Orthic Immature

NZ Soil Series Waimarama sandy loam Wanstead clay loamUS soil classification

Order Inceptisols InceptisolsSuborder Dystrudepts Eutrudepts

Olsen P (mg/kg) 25 16Anion storage Capacity (%) 43 21N fertiliser applied

(kgN/ha/yr)20 0

P fertiliser applied(kgP/ha/yr)

20 15

2 A livestock unit (SU) is the feed requirement used as the basis of comparisonfor different classes and species of stock. It expresses the annual feed requirements,equivalent to one 55 kg ewe rearing a single lamb. 1 SU requires approximately520 kg of good quality pasture dry matter per year.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 205

(version 6.0) (Wheeler et al., 2008) were used to generate therequired data.

2.2. General principles for quantifying and valuing ecosystemservices

The methodology used for the quantification and economicvaluation of ecosystem services follows the six guiding principlesfrom Dominati et al., (2014):

� The quantification of ecosystem services needs to be veryspecific to the benefits directly useful to humans, not to beconfused with the processes underlying ecosystem functioningand the supporting processes behind the formation and main-tenance of natural capital stocks.

� To determine how ecosystems provide ecosystem services, theproperties and processes at the origin of the provision of eachservice need to be described and quantified.

� Differentiate the contributions of the natural capital and addedor built capital (e.g. infrastructures, inputs such as fertilisers orirrigation water) to the flow of ecosystem services whendefining proxies to quantify each service. Identify where andhow external drivers such as climate and land use impact onthe provision of services through their impact on both naturalcapital stocks and processes.

� Analyse the impact of degradation processes, in this caseerosion, in addition to the effect of soil conservation measures,on natural capital stocks and thereby the flows of ecosystemservices.

� Base the economic valuation on the bio-physical measures ofthe services that are relevant for the chosen scale and land-use.

These principles align with the model described by DEFRA(2007) and used by Posthumus et al. (2013).

2.3. Quantification of the provision of ecosystem services for a sheepand beef operation

The quantification and economic valuation of ecosystem ser-vices was realised at the paddock scale (1 ha) for each of the twolandscape units (rolling and steep) of the hill-country sheep andbeef operation studied. For each service, a proxy to measure theservice was determined based on ecosystem services definitionsand Dominati et al. (2010) framework. The data needed tocalculate the proxy was generated from the OVERSEERs nutrientbudget model (version 6.0) (Wheeler et al., 2008) or sourced fromthe literature. Changes in the value of each proxy over time werebased on published research results when available, or assump-tions. The economic valuation was then based on the measure ofthe proxy for each service. Although, ‘precise’ economic values arepresented as results here, we recommend treating them as ‘ordersof magnitude’.

The detail of the steps in the quantification and economicvaluation for each of the provisioning and regulating ecosystemservices are presented in Table 2. Definitions of each of theprovisioning and regulating services are given by Dominati et al.(2010). The reader can also refer to Dominati et al., (2014) whichpresents in details the methodology applied to a dairy grazedsystem. In this study, where research knowledge and data werelacking, assumptions were made regarding changes to the mea-sure of each service. Assumptions are listed in Table 4. The culturalservices associated with a sheep and beef operation were notconsidered in this study.

2.4. Impact of shallow mass movement erosion on soil properties, theflow of ecosystem services and recovery

The main forms of erosion in New Zealand are (Cairns et al.,2001; Lynn et al., 2009):

� Mass movement erosion – which occurs when heavy rain orearthquakes cause whole slopes to slump, slip or landslide.Storms are the primary triggers. This is the most common formof erosion in the hill country.

� Fluvial erosion – which occurs when running water digsshallow channels or deeper gullies into the soil.

� Surface erosion – which occurs when wind, rain or frost detachsoil particles from the surface, allowing them to be washed orblown off the paddock (occurs largely outside the hill country.)

� Sediment erosion – activities involved in earthworks, planta-tion forests, cropland and pasture management may result insediment loads being mobilised and entering watercourses.

The erosion form addressed here which occurred in Hawke'sBay region in April 2011, is mass movement erosion. On unstableslopes, thousands of landslides can be triggered by high-magni-tude/low-frequency storms with estimated return periods exceed-ing 50 years (McIvor et al., 2011). In New Zealand, the distinction ismade between shallow (in the soil, easily stabilised) and deep (inunderlying regolith, where stabilisation is difficult) mass move-ments (Crozier, 1986; Cairns et al., 2001; Lynn et al., 2009).

Four forms of shallow mass movement are common in the NewZealand landscape:

� Slips – These are shallow landslides in soil or weatheredregolith on steep slopes or on low-angle slopes where regolithis susceptible to rupture.

� Earth flows – These are shallow flows on low-angle slopes,where regolith is susceptible to plastic deformation, or incolluvium (slip debris) on foot slopes.

� Debris avalanches—shallow, rapid landslides in regolith onupper mountain slopes.

� Debris flows – shallow, rapid flows in colluvium (avalanchedebris) on lower mountain slopes.

In this study no distinction was made between the differenttypes of shallow mass movements; therefore when the termlandslide is used, we refer to shallow mass movement erosion.Future studies could examine in more detail the impact differencetypes of erosion have on soil properties and recovery.

2.4.1. Loss of ecosystem services from erosionThe first step in quantifying the loss of ecosystem services from

an erosion scar and subsequent recovery was establishing the stateof the soil natural capital stocks after a shallow landslide. Existingresearch on the soil properties of erosion scars and soil recoveryafter landslides (Rosser and Ross, 2011a; Rosser and Ross, 2011b),as well as assumptions made by the authors, were used to describethe status of the natural capital stocks remaining after a shallowmass movement, and how that, and the provision of ecosystemservices from those natural capital stocks, recover in subsequentyears. Recovering services were re-quantified and re-valued usingthe methods presented in the previous section (Table 2).

The quantification and valuation of ecosystem services wasrecalculated for a paddock in a steep landscape unit immediatelyafter a shallow mass movement. It was assumed the soil displacedduring the shallow mass movement was deposited elsewhere, e.g.downslope covering existing pasture, and therefore still able toprovide services. It would be very interesting to consider the

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215206

Table 3Details of the additional ecosystem services provided by wide-spaced trees to a grazed pasture system, their quantification, proxy and economic valuation.

Services What to measure Parameters/indicators used toquantify the service

Formula ES measure Valuation method usedat the catchment scale

Feed QuantityTree

Tree growth Amount of foliage animals can eat,sustainable harvest (literature)

Amount of foliage animals can eat/year Equivalent inpasture drymatter

Pasture dry mattermarket price

Wood- Fibre Tree growth Amount of wood grown (literature) Amount of wood harvested Wood kg/ha after20–30 years

Wood market price

Provision ofshade toanimals

Impact of shade onanimal growth

Grazing time / day (literature) Growth with increased dry matterutilisation with shade�normal growth

kg meat/ha/yr Meat market price

Provision ofshelter toanimals

Impact of shelter onyoung animal survivalrates

Lamb and calf losses (literature) Number of young animals withshelter�normal number of younganimals

kg meat/ha/yr Meat market price

Net carbonaccumula-tion (tree)

Net C flows C stocks and variations (literature) Net C accumulation in wood t C/ha/yr Market prices of C

Table 2Detailed description, quantification, proxy and economic valuation of provisioning and regulating ecosystem services.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 207

provision of ecosystem services from such areas which potentiallypresent improved natural capital after the landslide. However, thiswas beyond the scope and boundaries of the present study. It wasassumed that immediately after the landslide (in the first year),soil would not grow pasture, support infrastructures or animals,accumulate carbon (C), or regulate greenhouse gases (GHG)(Rosser and Ross, 2011b). Therefore, the provision of those serviceswas assumed to be nil (Table 4).

After a shallow landslide water storage by the soil profile isreduced by 70% (Rosser and Ross, 2011b; Stavi and Lal, 2011).It was also assumed that the capacity of bare ground to filternutrients and contaminants decreased by 70%. On bare ground, itwas assumed than only 5% of the dung deposited was decomposedproperly. Since no pasture would grow on bare ground immedi-ately after a shallow landslide, the provision of regulation of pestand disease populations was assumed to be maximal (Table 4).

2.4.2. Recovery of the ecosystem services after erosionA number of studies have quantified the recovery rates of some

topsoil properties and pasture characteristics following massmovement erosion (Lambert et al., 1984; Sparling et al., 2003;Rosser and Ross, 2011b). This data was used to calculate therecovery of ecosystem services provision as the bare groundweathered and other pedological processes initiate soil develop-ment under a pastoral vegetation cover.

It has been shown that after a shallow landslide, pasturerecovery beyond 80% of uneroded level is unlikely even after50 years (Lambert et al., 1984; Rosser and Ross, 2011b). It wasassumed here that the provision of food quantity (pasture growth)and quality would recover linearly to 60% after 20 years, 80% after50 years and then plateau (Table 4).

The loss of depth of soil profile during a shallow landslide leadsto reduced drainage and water holding capacity. As mentionedearlier, some of the soil material would be deposited elsewhere,but this was not considered in this study. This means landslidesscars are likely to be wetter for extended periods limiting thesupport service. It was assumed that the support to animals wouldrecover up to 50% in 20 years and then further improvementswould be very slow. Similarly, flood mitigation and the filtering ofnutrients and contaminants are highly-dependent on the depth of

the soil profile and the nutrient status of the topsoil. It wasassumed that these services will recover up to 40% in 20 years,50% beyond 50 years, with further recovery tied to pedologicaltime scales. Microbiological activity in the new topsoil recovers upto 80% of the original topsoil in around 30 years (Sparling et al.,2003). To simulate the recovery of the decomposition of wastes,we assumed a recovery of 60% after 20 years, and 80% after 50years (Table 4).

The new top soil being formed accumulates C faster thanuneroded soils (Lambert et al., 1984; Sparling et al., 2003; Page etal., 2004) even though total C levels may never recover tothe uneroded top soil levels within human lifetimes (Sparlinget al., 2003; Rosser and Ross, 2011b). It was assumed that net Caccumulation rates were 10 times faster than uneroded levels in thefirst 10 years following a shallow landsliding, decreasing overtime,leading to total C levels in the new topsoil reaching 80% of unerodedlevels after 45 years (Sparling et al., 2003; Page et al., 2004). Theregulation of greenhouse gas (GHG) emissions also depends ondrainage and nutrients status, so it was therefore assumed that thisservice will recover up to 50% in 20 years (Table 4).

As pasture re-establishes on the landslide scar, and topsoilstarts forming and accumulating, pests also return. It was assumedthat, like a newly sown pasture, initial infestation rate will be highbetween year 2 and 5 after the landslide, before declining to thelevels found on an uneroded site.

2.5. Impact of soil conservation practices on the flow of ecosystemservices:

Options for reducing the risk of shallow mass movements inhills country usually are (Crozier, 1986; Cairns et al., 2001; Lynnet al., 2009)

� Spaced planting of trees in pasture, to provide the soil with rootreinforcement, or

� Land use change from grazing to commercial timber trees, or� Reversion to native scrub cover.

The first option is examined in this study. The ability ofwide-spaced tree planting to reduce the occurrence of shallowlandslides is well understood. Douglas et al., (2011) showed that

Table 4Assumptions made for the base analysis and sensitivity analysis.

Ecosystem services Soil recovery–baseassumptions

Soil recovery–sensitivityanalysis

Wide spaced trees-baseassumption

Wide spaced trees-sensitivityanalysis

(% of pasture baseline) (% of pasture baseline) (% of pasture baseline) (% of pasture baseline)(% of initial assumption)

Years following erosion Years following erosion Age of trees Age of trees

0 20 50 0 20 50 10 years 20 years 10 years 20 years

Food quantity pasture (%) 0 60 80 0 30 40 85 75 70 50Food quantity tree (%) NA NA NA NA NA NA 50 50Food quality-pasture (%) 0 60 80 0 30 40 100 100 100 100Wood-fibre (%) NA NA NA NA NA NA 50 50Provision of support for human infrastructures (%) 0 0 0 0 0 0 0 0 0Provision of support for farm animals (%) 0 50 50 0 25 25 100 100 100 100Provision of shade (%) NA NA NA NA NA NA 50 50Provision of shelter (%) NA NA NA NA NA NA 50 50Flood mitigation (%) 30 40 50 15 20 25 110 130 100 100Filtering of nutrients and contaminants (%) 30 40 50 15 20 25 110 130 100 100Decomposition of wastes (%) 5 60 80 3 30 40 105 110 100 100Net carbon accumulation (soil) (%) 0 1000 83 0 500 42 60 60 30 30Net carbon accumulation (tree) (%) NA NA NA NA NA NA 50 50Nitrous oxide regulation (%) 0 40 50 0 20 25 240 240 100 100Methane oxidation (%) 0 40 50 0 20 25 200 200 100 100Regulation of pest and disease populations (%) 113 100 100 50 50 50 100 100 100 100

NA: Not applicable

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215208

wide-spaced trees can reduce landslide occurrence by 95%,compared to pasture only landscapes. It was assumed in theexamination of the benefits of wide-spaced tree planting (50stems per hectare, 30% canopy cover) on the provision of ecosys-tem services from soils under a pastoral use, that once the treesare planted, no more landslides occur for the next 20 years, on100% of the area covered. In reality the trees will have limitedutility against erosion until 5–10 years of age. Additional servicesfrom trees, that would otherwise not be available in a grasslandsystem, were included in the analysis (Table 3).

2.5.1. Changes to ecosystem services provision under treesData from the literature were used to quantify the influence the

wide-spaced planted trees had on the provision of ecosystemservices from the pasture soil. The following assumptions weremade. Wide-spaced trees reduce pasture production by 15% whentrees are aged 5–10 years and by 25% when trees are aged 10–20years and over (Guevara-Escobar et al., 2007; McIvor and Douglas,2012). The trees have no influence on pasture quality (Table 4).

Trees intercept 5–35% of rainfall (Guevara-Escobar et al., 1998;Benavides et al., 2009) depending on tree density and crown size,therefore we assumed that runoff was decreased by 10% underyoung trees (o10 years) and 30% under older trees (420 years).

Increased water uptake by pasture in combination with treesand improved soil physical characteristics (Benavides et al., 2009)were assumed to improve the soil's filtering capacity by 10% underyoung trees (o10 years) and 30% under older trees (420 years)(Table 4).

Increase in carbon (C) and nitrogen (N) mineralization (Benavideset al., 2009) and increased soil pH and exchangeable cations (Ca, K andMg) (Guevara-Escobar et al., 2002) were identified under wide-spacedtrees compared to open pastures. Therefore we assumed the decom-position of wastes was 5% greater under young trees (o10 years) and10% greater under older trees (420 years).

Soil C and N content (top 10 cm) are greater in open pasture thanin a poplar-pasture system (Guevara-Escobar et al., 2002; Benavideset al., 2009); therefore the net C accumulated in soils by a pasture-treesystem was assumed to be 60% of open pasture (Table 4).

Since pasture-tree systems are usually better drained (Benavideset al., 2009) it was assumed that N2O emissions would be 50% lowerand CH4 oxidation would double compared to open pasture.

The level of regulation of soil pest and diseases populationwas assumed to be the same for open pasture or under poplars(Table 4).

2.5.2. Additional ecosystem services from treesConservation trees are an ecological infrastructure investment

(Jury et al., 2011) that provides additional support and resilience tothe pastoral ecosystem. Both McGregor et al. (1999) and Parminteret al. (2001) identified several benefits from introducing wide-spaced-planted trees into a pasture system, including shelter fromextreme events for farm animals, shade throughout the year, soilstability, vista, food source for native birds, wood fibre and analternate forage source for grazing animals during summer dryperiods, which coincides with mating and hence provides insur-ance against decline in ovulation rates (Orsborn et al., 2003). Theseservices were added to the list above (Table 3).

Pollarding willow and poplar in the summer months canprovide a supplementary fodder source in summer-dry hill coun-try (Douglas et al., 2006). It was assumed here that trees could beused for forage when aged 5 years or older. The amount of forageproduced was calculated for 10 and 20 year old trees assumingthey have not been pollarded previously. The value of the availableforage was added as a service. Timber is another service that treesprovide as they can be harvested for their wood. The provision of

shelter and shade for animals is an emerging reason for greatertree-planting on pastoral farms. Trees are an option for reducingthe risk of stock losses from extreme climatic events and moder-ating extended weather extremes (e.g. temperature) to protectcapital stock and sustain animal growth rates. The availability ofshade has been shown to prolong animal grazing each day(Betteridge et al., 2012), so an increased dry matter utilisationwas used to measure these services. Shelter has the potential tomake a significant difference to lamb and calves survival rates andinitial growth rates in the spring months, so it was assumed thatlamb and calf losses were reduced by 25% between scanning andweaning (Parminter et al., 2001) for trees older than 5 years and50% for trees older than 10 years. Established soil conservationtrees provide an opportunity to claim carbon (C) credits under theNew Zealand Emissions Trading Scheme. Net C accumulated inpoplar trees was quantified (Table 3). The influence of a tree-pasture system on the GHG regulation and the regulation ofabove-ground pest and disease populations of the system is stillpoorly understood so we assumed here that there were noimmediate additional benefits other than the ones considered insoils in the section above.

2.6. Sensitivity analysis

A sensitivity analysis was included to investigate the influencethe assumptions made in the quantification of the ecosystemservices and calculation of economic value of the services had onthe results and subsequent interpretations. Two separate sensitiv-ity analysis were conducted.

1. The recovery of the ecosystem services after an erosion eventwas assumed to be only half of the original recovery rates(Table 4). The implications of this to the cost of erosion werethen recalculated.

2. The wide-spaced trees only grow at half the rate and produceonly half of the additional ecosystem services (Table 4). Theimplications of the benefit cost analysis of a soil conversationinvestment were also recalculated.

2.7. Economic and financial analysis

To determine the economic value of each service at the farmscale neo-classical economic valuation techniques, including mar-ket prices, defensive expenditures, replacement cost and provisioncost, were used (Tables 2 and 3). A mixture of valuation methodshad to be used as many services, such as regulating services, arenot traded in markets and therefore lack direct market values.

Basing the economic valuation of each service on its direct bio-physical measure, as opposed to looking at costs foregone with thesoil conservation measure (Posthumus et al., 2013) enables a directlink to be made between soil properties, soil change and a changein the value of the service.

When costs of infrastructure were used as proxies for the valueof the service (defensive expenditures, replacement cost andprovision cost), the construction costs of the infrastructure wereannualised over its lifetime using a depreciation rate of 10% andadded to annual maintenance costs to determine the annual valueof the service.

According to the information presented above on changes tothe systems, the value of each service was calculated for a rollinglandscape unit uneroded under pasture, a steep landscape unituneroded under pasture, a steep landscape unit eroded immedi-ately after a shallow landslide (year 0), and 20 and 50 years after,and finally a steep landscape unit with a pasture-tree system with10 and 20 year old trees (Table 5).

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 209

To investigate the impact of an investment in soil conservationon the provision of ecosystem services, the flows of services wereconsidered on a per hectare basis over 20 years, using threescenarios illustrated in Fig. 1

� Scenario A – Business as usual: provision of ecosystem servicesfrom a sheep and beef farm for the two landscape units with nolandslides and no addition of conservation trees,

� Scenario B – Shallow landslide followed by recovery: provisionof ecosystem services from the steep landscape unit of a sheepand beef farm with a landslide in year 0 and subsequentrecovery of soil and ecosystem services over 20 years,

� Scenario C – Soil conservation: Planting of conservation treesat 50 stems per hectare on the steep landscape unit at year 0to reduce the risk of soil erosion, and subsequent tree devel-opment over 20 years.

For each of these scenarios the Present Value (PV) of the flow ofecosystem services (benefits) was calculated over 20 years. Thepresent value (PV) of cash flows is a widely used criterion inbenefit-cost analysis. It calculates the present value of a sum ofmoney in year t (here the economic value of ecosystem services inyear t) by discounting it at the rate r, arising between the presentand a future date. The PV of a sum of money received in the future

is calculated using the following equation:

PV ¼ Vt

ð1þrÞt

where PV is the present value, t is the year, Vt is the value of thecash flow in year t, and r is the discount rate.

Different discount rates, between 0% and 10%, were used in thisstudy to show the sensitivity of the analysis to the choice ofdiscount rate.

The flow of services was assumed to be constant for 20 yearsfor the two uneroded units, rolling and steep. For the steep unitwith shallow landslide followed by recovery, it was assumed thatthe value of the services was recovering linearly from the valuecalculated for immediately after the landslide to the value after20 years recovery (Table 5). For the soil conservation on steeplandscape, it was assumed that the value of the services wassimilar to steep land with pasture only in the first 5 years, similarto the value calculated for a tree-pasture system in year 10between year 6 and 10, and similar to the value calculated for atree-pasture system in year 20 between years 10 and 20 (Table 5).

Finally a benefit-cost (BCA) analysis of an investment in soilconservation on the steep landscape unit was realised. The extracosts and benefits associated with conservation trees were con-sidered over 20 years on a per hectare basis, including the treemanagement costs and differences in the value of the ecosystem

Table 5Economic value (NZD/ha/yr) of ecosystem services from uneroded pasture on rolling and steep landscape units, eroded steep land 0, 20 and 50 years after a shallow landslideand on a steep land pasture planted with wide-spaced trees.

Soil service Uneroded Eroded Steep landscape unit Steep landscape unit with trees

Rolling landscape unit Steep landscape unit 0 years 20 years 50 years 10 years 20 years harvested

Food quantity pasture 745 484 0 290 387 411 363Food quantity tree NA NA NA NA NA 105 210Food quality-pasture 29 29 0 17 23 29 29Wood-fibre NA NA NA NA NA 104 104Provision of support for human infrastructures 204 0 0 0 0 0 0Provision of support for farm animals 53 33 0 17 17 33 33Provision of shade NA NA NA NA NA 58 58Provision of shelter NA NA NA NA NA 9 19Flood mitigation 1155 911 273 364 456 938 990Filtering of nutrients and contaminants 2227 1800 634 807 978 1847 1940Decomposition of wastes 336 127 21 76 102 149 170Net carbon accumulation (soil) 6 2.2 0 22 1.8 1.3 1.3Net carbon accumulation (tree) NA NA NA NA NA 150 300Nitrous oxide regulation 2 1.2 0 0.5 0.6 2.7 2.7Methane oxidation 0.08 0.08 0 0.03 0.04 0.06 0.06Regulation of pest and disease populations 328 328 371 328 328 328 328Combined value (NZD/ha/yr) 5085 3717 1299 1921 2293 4165 4548

NA: Not applicable

Fig. 1. Three scenarios – Scenario A – typical east coast hill country sheep and beef grazing; scenario B – Shallow mass movement erosion followed by recovery and scenarioC – Soil conservation in hill country. (B-photo Brenda Rosser; C-photo Grant Douglas)

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215210

services provided between a normal pasture and a tree-pasturesystem. The Net Present Value (NPV) of the investment on a perhectare basis was calculated by discounting the streams of benefitsand costs at the rate r arising between the present and a futuredate, using the following formula:

NPV ¼ ∑t

t ¼ 0

Bt�Ct

ð1þrÞt

Where NPV is the net present value (PV of benefits � PV ofcosts), t is the year (0–20), Bt is the value of the benefits in year t, Ctis the value of the costs in year t, and r is the discount rate.

Four options were considered for the steep landscape unit as acombination of selling the trees for wood and considering or notthe value of the ecosystem services provided in the BCA

� Option 1: Trees are not sold for wood and the value of theecosystem services provided are not considered in the BCA.

� Option 2: Trees are sold for wood and the value of the eco-system services provided are not considered in the BCA.

� Option 3: Trees are not sold for wood, but the value of theecosystem services provided are included in the BCA.

� Option 4: Trees are sold for wood and the value of theecosystem services provided are included in the BCA.

For each option, the NPV of the investment was assessed usingfour discount rates between 0% and 10%.

3. Results and discussion

3.1. Economic Value of the provision of ecosystem services

The economic value of ecosystem services (as calculated accord-ing to Tables 2 and 3) provided by uneroded pasture grazed bysheep and cattle was NZD 5085 ha�1 yr�1 for the rolling landscapeunit and NZD 3717 ha�1 yr�1 for the steep landscape unit (Table 5).Of the eleven ecosystem services, filtering of nutrients and con-taminants had the greatest economic value (44–49%), followed byflood mitigation (23–25%) and then the provision of food (13–14%).Regulating services, which are usually not traded in markets orconsidered in decision making, had an economic value four-timesthat of the provisioning services for the rolling landscape unit andalmost six-times greater for the steep landscape unit (Fig. 2).

The economic value of ecosystem services provided bybare ground following a shallow landslide on steep land was NZD1299 ha�1 yr�1, a decrease of 65% of the services from unerodedland (NZD 3717 ha�1 yr�1), based on the assumptions in Table 4.

Based on the work of Rosser and Ross (2011b) and theassumptions made, the initial recovery of provisioning servicesfollowing a shallow landslide on the steep landscape unit wasrelatively quick with 59% recovery in the first 20 years, and after50 years up to 78% of the economic value of the services of theuneroded landscape (Fig. 2). Further recovery is linked to long-term pedological processes. Recovery of the regulating serviceswas predicted to be much slower, with only a 50 and 59% recoveryof these services after 20 years and 50 years, respectively (Fig. 2).In the first few years after the erosion event, soil C accumulationand pest regulation rates are high, before slowing and stabilising.

Assuming the recovery of the soil on the eroded land was onlyhalf of what was used in the initial analysis, the economic value ofecosystem services would drop to 53% (NZD 1013 ha�1 yr�1) and53% (NZD 1205 ha�1 yr�1) of the value using initial assumptionsfor the steep land recovering after 20 and 50 years respectively(Fig. 3).

The long-time frames required for the recovery of regulatingservices after an erosion event (Fig. 2) are generally given little

thought and as a consequence do not feature in most assessmentsof land degradation following erosion. Until the impacts of erosionon these services have a recognised value, the long-term con-sequences of erosion on human well-being will continue to beunder-estimated.

Planting wide-spaced trees on pasture on the steep landscapeunit, while decreasing the provision of specific services (Table 5), ingeneral increased the provision of baseline ecosystem services andadded extra services including forage from trees, wood, provision ofshade and shelter for animals and net C accumulation in wood(Fig. 2). The presence of soil conservation trees increased theeconomic value of the ecosystem services (as calculated accordingto Tables 2 and 3) provided from NZD 3717 ha�1 yr�1 to NZD4165 ha�1 yr�1 after 10 years and NZD 4548 ha�1 yr�1 after 20years (Table 5), an increase in value of 12% and 22%, respectively.

Had the conservation trees only grown at half the rate andassociated services also halved, the economic value of eco-system services would be 91% (NZD 3781 ha�1 yr�1) and 84%(NZD 3817 ha�1 yr�1) of the value using initial assumptions forthe steep land with 10 and 20 year old trees, respectively (Fig. 3).

While not included in the analysis, conservation trees plantedon-farm will impact on the provision of cultural services, such asvista and landscape aesthetic values (Swaffield and McWilliam,2014), land prices, and sense of stewardship of the land. Further,conservation trees also impact on habitat. For example poplar isknown to be a food source for some native bird species in earlyspring (Yao and Kaval, 2010). Inclusion of these services would addto the economic value of the ecosystem services of a sheep andbeef operation which has soil conservation plantings as part of thefarms ecological infrastructure.

3.2. Net present value of the flow of ecosystem services

The three scenarios Scenario A – Business as usual, ScenarioB – Shallow landslide followed by recovery, and Scenario C – Soilconservation were considered over 20 years to explore the NetPresent Value of an investment in soil conservation. For each ofthese scenarios the NPV of the flow of ecosystem services, that isthe flow of benefits coming from the land, was calculated over 20years, using initial assumptions and different discount rates. Whenconsidering the recovery of the steep landscape unit after a land-slide (100% eroded), it was assumed that the economic value of theecosystem services provided was NZD 1299 ha�1 yr�1 in the firstyear (economic value of the services after erosion), increasinglinearly to NZD 1922 ha�1 yr�1 over 20 years (economic value ofservices after 20 years recovery) (Table 5). When considering theprovision of ecosystem services from the steep landscape unitunder wide-spaced trees, it was assumed that the provision ofecosystem services was similar to pasture in the first 5 years (NZD3717 ha�1 yr�1) (Table 5). Then, the provision of ecosystem serviceswas assumed to be worth NZD 4165 ha�1 yr�1 between year 6 and10, and NZD 4548 ha�1 yr�1 between years 10 and 20 (Table 5).

The NPV of the ecosystem services from the steep landscape unit,for each of the three scenarios, are presented in Fig. 4, with differentdiscount rates. When calculating the NPV of the flow of ecosystemservices over 20 years for the three scenarios, changing the discountrate from 10% to 0% more than doubles the NPV (Fig. 4). Trees plantedon the steep landscape unit increased the NPV of the ecosystemservices by around 10% (using initial assumptions) but most impor-tantly, by preventing soil erosion, reduces the risk of a loss of NPV ofthe flow of ecosystem services of around 58% over 20 years.Comparison of the provision of ecosystem services from the 20 yearwide-spaced-planted tree-pasture system with the open pasture 50years on from a landslide (Table 5), serves to highlight the key role oftrees in protecting soil natural capital stocks and securing keyregulating services.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 211

After the sensitivity analysis, the NPV of ecosystem serviceswas around 53% and 91% of the initial values presented in Fig. 4,for Scenario B and Scenario C respectively.

After a single landslide, the economic value of the ecosystemservices provided by the landscape under a sheep and beef oper-ation decreased by 64% from NZD 3717 ha�1 yr�1 to NZD1299 ha�1 yr�1 (Table 5). Assuming the land recovers over 20years without another landslide, this represents a loss of NPV ofNZD 33,989/ha (from NZD 59,020 to NZD 25,031 ha�1) over 20years using a 3% discount rate (Fig. 4). Application of this loss inNPV across the 4300 ha affected by the April 2011 storm equates toNZD 146 million. This is an indication of the NPV of the ecosystemservices permanently lost from the storm event, including the lossof pasture production and income to the farmer. This must beadded to the actual material costs of the storm, estimated at NZD39 million for infrastructure, and land, personal and commercialdamage claims. This gives a total cost of the April 2011 storm inexcess of NZD 185 million or more than four times the materialclean-up cost that is quoted as the storm cost incurred by thecommunity.

Soil conservation trees not only reduce the risk of erosion conser-ving the equivalent of NZD 33,989 ha�1 of NPV (Fig. 4), but also addextra services that when valued add an extra NZD 7043 ha�1, fromNZD 59,020 ha�1 for uneroded steep pasture to NZD 66,063 ha�1 forsteep pasture with trees (using a 3% discount rate) (Fig. 4). If half of the4300 ha affected by the storm on the east coast on April 2011 hadbeen planted in trees, reducing the risk of erosion, the added NPV ofthe ecosystem services provided by that land would have amount toNZD 15.1 million over 20 years using a discount rate of 3%.

Further, those wide-spaced trees, by preventing erosion, wouldpotentially have preserved a value of NZD 108 million (NZD 73million of NPV of ecosystem services not lost, NZD 15.6 million ofNPV of potential extra ecosystem services from trees, and NZD 20million in costs that could have been avoided), for an originalinvestment of NZD 1.6 million in planting 2150 ha, based on costs ofNZD 736 ha�1 to plant the trees at 50 sph. In this case, every NZD 1spent on soil conservation trees is worth NZD 68 of NPV of avoidedinfrastructure costs and lost services, and extra services provided.

These calculations put the cost of soil erosion and the value ofsoil conservation in a different light from conventional evaluation

Fig. 2. Economic value (NZD/ha/yr) of ecosystem services provided by pasture grazed by sheep and cattle, on uneroded rolling and steep land, steep land immediately after ashallow landslide and following 20 and 50 years of recovery, and steep land planted with 10 and 20 year old wide spaced trees.

Fig. 3. Combined economic value (NZD/ha/yr) of ecosystem services provided by pasture grazed by sheep and cattle, on uneroded rolling and steep land, steep landimmediately after a shallow landslide and following 20 and 50 years of initial and half initial assumed recovery rates, and steep land planted with 10 and 20 year old widespaced planted trees growing at the initial or half initial growth rate.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215212

of soil conservation practices. Accepting a number of the assump-tions in the quantification and valuation of the services are open tointerpretation and can be challenged, the methodology providesinsights into the long-term effects of land degradation on theprovision of the whole range of ecosystem services and theincreased resilience brought with the addition of conservationtrees on land at risk of erosion in storm events. This informationshould be available to decision makers for informing policydevelopment and investment in soil conservation practices.

3.3. Soil conservation benefits and benefit-cost analysis

The benefit-cost analysis of an investment in soil conservationexplored four options that included a combination of selling the

trees for wood and considering or not the economic value of theecosystem services (as calculated according to Tables 2 and 3, andreported in Table 5). For each option, the NPV of all the extrabenefits and costs associated with conservation trees discountedover 20 years was calculated using four discount rates between 0%and 10% (Fig. 5).

The economic value of the additional ecosystem servicesprovided by the wide-spaced trees (Table 5), beyond the serviceprovision from the open pasture, is not usually considered inBenefit cost analysis of soil conservation practices. Here, all thecosts associated with growing poplars, including planting, prun-ing, pollarding and harvesting were considered, alongside theadditional services this ecological infrastructure investmentprovided.

Fig. 4. Net present value (NPV; NZD/ha) over 20 years of the flow of ecosystem services associated with uneroded steep land (Scenario A), eroded steep land recovering(Scenario B) and steep land planted with conservation trees (Scenario C), for different discount rates.

Fig. 5. Net present value (NPV; NZD/ha) over 20 years of four different management options at for four discount rates.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 213

The option of selling the trees for timber after 20 years was alsoconsidered despite this option having limited applicability atpresent. The costs associated with tree planting were assumed tobe NZD 736 ha�1 (for 50 stems per hectare) as plantation costsand NZD 200 ha�1 in year 7 and 17 for pruning. The cost ofharvesting and transport was subtracted from the net revenuecoming from the timber sale in year 20. The NPV of the differentoptions are presented in Fig. 5.

Planting soil conservation trees without harvesting them(Option 1) is not profitable if the extra provision of ecosystemservices is not considered (NPV¼�NZD 998 at 3%) (Fig. 5).However, if the extra provision of ecosystem services is considered(Option 3) the NPV of the investment is positive regardless of thediscount rate used (Fig. 5). If the trees are harvested for timberafter 20 years but the ecosystem services not considered (Option2), the investment is only profitable for discount rates below 5%.Again, if considering the economic value of the extra provision ofecosystem services from trees as well as timber sale (option 4), theNPV of the investment is greatly positive regardless of thediscount rate (Fig. 5).

The sensitivity analysis showed that with lower tree growthrates, the economic value of the ecosystem services provided bythe steep landscape unit with conservation trees is reduced, andthe investment is only profitable if the trees are harvested forwood. With slower growth rates it would be necessary to extendthe analysis out beyond 20 years to assess the merits of a spaceplanted trees versus other soil conservation investment options.Ultimately, the final decision on a soil conservation investmentwill be dependent on the type, risk and severity of erosion, thetime period over which the investment is investigated, in addi-tion to the decision to include the impact on ecosystem servicesand the economic value associated with each service. Further, theallocation of ecosystem services economic values to privateversus public beneficiaries and the level of the return on thenext best use of the money invested in soil conservation will alsobe factors. This study shows how much of a difference inclusionof the whole range of ecosystem services and their economicvalue makes to the estimated NPV of conservation practices.

The ecosystem services approach described here offers a morecomplete picture of the “costs” of soil erosion, its long-term effectson the provision of services, and the ‘value’ of soil conservationpractices (Robinson et al., 2014).

4. Conclusion

This study through a live case on the East Coast of the NorthIsland of New Zealand addresses soil conservation and shows howan ecosystem services approach can be used to quantify and valuethe long-term costs of soil erosion, here shallow mass movements,and the multiple benefits from soil conservation.

While this study is based on current research knowledge, someassumptions were made where data was lacking regardingchanges to the measure of ecosystem service from eroded land-scapes and landscapes with soil conservation trees. When databecomes available this study can be revisited.

The methodology tested in this study demonstrates that usingexisting and detailed information about ecosystems change at asmall scale to quantify the provision of provisioning and regulatingecosystem services from agro-ecosystems is not only feasible,but also extremely informative for decision makers for not onlypolicy development, but also for evaluating the impacts of policy.Understanding how current and future investments in ecologicalinfrastructure, such as soil conservation trees, are likely to changethe flow of ecosystem services from managed landscapes is criticalto assess the efficiency, cost-effectiveness and sustainability of

resource management policies, and to increase political and publicawareness of the value of land and the long term costs of landdegradation.

Acknowledgement

Funding for this study was provided by the Envirolink toolsprograms through Hawke's Bay Regional Council, as well as theRutherford foundation of the Royal Society of New Zealand and theSustainable Land Use Research Initiative.

Thank you to colleagues and reviewers for their valuablecomments and suggestions.

References

Basher, L.R., Botha, N., Dodd, M.B., Douglas, G.B., Lynn, I., Marden, M., McIvor, I.R.,Smith, W., 2008. Hill country erosion: a review of knowledge on erosionprocesses, mitigation options, social learning and their long-term effectivenessin the management of hill country erosion. In: Phillips, C., Collins, A. (Eds.),Landcare Research Contract Report, vol. 0708/081; 2008, p. 201 (pp).

Benavides, R., Douglas, G.B., Osoro, K., 2009. Silvopastoralism in New Zealand:review of effects of evergreen and deciduous trees on pasture dynamics.Agrofor. Syst. 76, 327–350.

Betteridge, K., Costall, D.A., Martin, S., Reidy, B., Stead, A., Millner, I., 2012. Impact ofshade trees on Angus cow behaviour and physiology in summer dry hillcountry: grazing activity, skin temperature and nutrient transfer issues. Pp(Occasional Report no. 25). In: Currie, L.D., Christensen, C.L. (Eds.), AdvancedNutrient Management: Gains from the Past — Goals for the Future. Fertilizerand Lime Research Centre, Massey University, Palmerston North.

Cairns, I., Handyside, B., Harris, M., Lambreschtsen, N., Ngapo, N., 2001. SoilConservation Technical Handbook. The Ministry for the Environment, Well-ington, New Zealand.

Crozier, M.J., 1986. Landslides: Causes, Consequences & Environment. Croom HelmLtd., London, Sydney.

DEFRA 2007: An Introductory Guide to Valuing Ecosystem Services. In: ed. London,Department for Environment, Food and Rural Affairs, ⟨http://ec.europa.eu/environment/nature/biodiversity/economics/pdf/valuing_ecosystems.pdf⟩.

Dominati, E.J., Mackay, A.D., Green, S., Patterson, M.G., 2014. A ‘soil change-based’methodology for the quantification and valuation of ecosystem services fromagro-ecosystems: a case study of pastoral agriculture in New Zealand. Ecol.Econ. 100, 119–129.

Dominati, E.J., Patterson, M.G., Mackay, A.D., 2010. A framework for classifying andquantifying the natural capital and ecosystem services of soils. Ecol. Econ. 69,1858–1868.

Douglas, G.B., McIvor, I.R., Manderson, A.K., Koolaard, J.P., Todd, M., Braaksma, S.,Gray, R.A., 2011. Reducing shallow landslide occurrence in pastoral hill countryusing wide-spaced trees. Land Degrad. Dev. 24, 103–114.

Douglas, G.B., Walcroft, A.S., Hurst, S.E., Potter, J.F., Foote, A.G., Fung, L.E., Edwards,W.R.N., van den Dijssel, C., 2006. Interactions between widely spaced youngpoplars (Populus spp.) and introduced pasture mixtures. Agrofor. Syst. 66,165–178.

Eastwood, C.; Krausse, M.; Alexander, R.R. 2001: Muddied waters: Estimating thenational economic cost of soil erosion and sedimentation in New Zealand. In:ed. Palmerston North, New Zealand, Landcare Research.

Guevara-Escobar, A., Edwards, A., Morton, R.N.W., Kemp, P.D., Mackay, A.D., 1998.Tree water use and rainfall partitioning in a mature poplar-pasture system. TreePhysiol. 20, 97–106.

Guevara-Escobar, A., Kemps, P.D., Mackay, A.D., Hodgson, J., 2002. Soil properties ofa widely spaced, planted poplar (populus deltoides) – pasture system in a hillenvironment. Aust. J. Soil Res. 40, 873–886.

Guevara-Escobar, A., Kemps, P.D., Mackay, A.D., Hodgson, J., 2007. Pasture produc-tion and composition under poplar in a hill environment in New Zealand.Agrofor. Syst. 69, 199–213.

Jones, K.E.; Levick, S.R.; Page, M.J. 2011. In: Processing and classifying satelliteimagery to assess the April 2011 storm induced landsliding in Hawke's Bay. In:ed. GNS Science 24 (pp).

Jury, W.A., Or, D., Pachepsky, Y., Vereecken, H., Hopmans, J.W., Ahuja, L.R., Clothier,B.E., Bristow, K.L., Kluitenberg, G.J., Moldrup, P., Simunek, J., Van Genuchten, M.T., Horton, R., 2011. Kirkham’s legacy and contemporary challenges in soilphysics research. Soil Sci. Soc. Am. J. 75, 1589–1601.

Lambert, M.G., Trustrum, N.A., Costall, D.A., 1984. Effect of soil slip erosion onseasonally dry Wairarapa hill pastures. N. Z. J. Agric. Res. 27, 57–64.

Lynn, I.; Manderson, A.; Page, M.; Harmsworth, G.; Eyles, G.; Douglas, G.; Mackay,A.; Newsome, P. 2009: Land Use Capability Survey Hand-book — a New Zealandhandbook for the classification of land — 3rd ed. Hamilton, Agresearch, Lincoln,Landcare Research, Lower Hutt, GNS Science.

McBratney, A., Field, D.J., Koch, A., 2014. The dimensions of soil security. Geoderma213, 203–213.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215214

McGregor, E., Mackay, A.D., Dodd, M.B., Kemp, P.D., 1999. Silvopastoralism usingtended poplars on New Zealand hill country: The opportunities. Proc. N. Z.Grassl. Assoc. 61, 85–89.

McIvor, I., Douglas, G.B., 2012. Poplars and willows in hill country – stabilising soilsand storing carbon. (Occasional Report no. 25). In: Currie, L.D., Christensen, C.L.(Eds.), Advanced Nutrient Management: Gains from the Past – Goals for theFuture. Massey University, Palmerston North, New Zealand (Fertilizer and LimeResearch Centre).

McIvor, I.; Douglas, G.B.; Dymond, J.R.; Eyles, G.O.; Marden, M. 2011: Pastoral HillSlope Erosion in New Zealand and the Role of Poplar and Willow Trees in ItsReduction. Soil Erosion Issues in Agriculture. Godone, D.; Stanchi, S. Eds.Available from: ⟨http://www.intechopen.com/articles/show/title/pastoral-hill-slope-erosion-in-new-zealand-and-the-role-of-poplar-and-willow-trees-in-its-reduction⟩, InTech.

MEA, 2005. Millennium Ecosystem Assessment: Ecosystems and Human Well-being: Synthesis. Island Press, Washington DC.

MfE 2007. Environment New Zealand 2007: State of the environment report.Wellington, New Zealand, Ministry for the Environment.

MPI 2012: Farm Monitoring Report 2012 – Pastoral Monitoring: Hawke's Bay/Wairarapa, sheep and beef In: ed. Wellington, New Zealand, Ministry forPrimary Industries.

Orsborn, S.; Gawith, P.W.; Cameron, D.J. 2003. Cost-benefits of supplementing eweswith willow and poplar foliage on a model hill country farm in Wairarapa.Using Trees on Farms. Grassland Research and Practice Series No.10. Charlton,J. F. L. Ed. Wellington, New Zealand, New Zealand Grassland Association.pp. 35–40.

Page, M.J., Trustrum, N., Brackley, H., Baisden, T., 2004. Erosion-related soil carbonfluxes in a pastoral steepland catchment, New Zealand. Agric. Ecosyst. Environ.103, 561–579.

Parminter, I., Dodd, M.B., Mackay, A.D., 2001. Economic analysis of poplar plantingon steep hill country. Proc. N. Z. Grassl. Assoc. 63, 127–130.

Posthumus, H.; Deeks, L.K.; Rickson, R.J.; Quinton, J.N.. 2013: Costs and benefits oferosion control measures in the UK. Soil use and management.

Robinson, D.A., Fraser, I., Dominati, E.J., Davídsdottir, B., Jónsson, J.O.G., Jones, L.,Jones, S.B., Tuller, M., Lebron, I., Bristow, K.L., Souza, D.M., Banwart, S., Clothier,B., 2014. On the Value of Soil Resources in the Context of Natural Capital andEcosystem Service Delivery. Soil Sci. Soc. Am. J. 78, 685–700.

Rosser, B., Ross, C.W., 2011a. Erosion soil loss and recovery on eastern north islandhillcountry – implications for nutrient management and pasture productivity.In: Currie, L.D., Christensen, C.L. (Eds.), Adding to the knowledge base for thenutrient manager. Massey University, Palmerston North, New Zealand (Occa-sional Report no. 24. Fertilizer and Lime Research Centre).

Rosser, B.J., Ross, C.W., 2011b. Recovery of pasture production and soil properties onsoil slip scars in erodible siltstone hill country, Wairarapa, New Zealand. N. Z. J.Agric. Res. 54, 23–44.

Sparling, G., Ross, D., Trustrum, N., Arnold, G., West, A., Speir, T., Schipper, L., 2003.Recovery of topsoil characteristics after landslip erosion in dry hill country ofNew Zealand, and a test of the space-for-time hypothesis. Soil Biol. Biochem.35, 1575–1586.

Stavi, I., Lal, R., 2011. Loss of soil resources from water-eroded versus unerodedcropland sites under simulated rainfall. Soil Use Manag. 27, 69–76.

Swaffield, S.R., McWilliam, W.J., 2014. Landscape aesthetic experience and ecosys-tem services. In: Dymond, J.R. (Ed.), Ecosystem services in New Zealand —

conditions and trends. Manaaki Whenua Press, Lincoln, New Zealand.Wheeler, D.M., Ledgard, S.F., de Klein, C.A.M., 2008. Using the OVERSEER nutrient

model to estimate on-farm greenhouse gas emissions. Aust. J. Exp. Agric. 48,99–103.

Yao, R.T., Kaval, P., 2010. Valuing biodiversity enhancement in New Zealand. Int. J.Ecol. Econ. Stat. 16, 26–42.

E.J. Dominati et al. / Ecosystem Services 9 (2014) 204–215 215