SOIL 4 259ndash266 2018httpsdoiorg105194soil-4-259-2018copy Author(s) 2018 This work is distributed underthe Creative Commons Attribution 40 License

SOIL

Spatial assessments of soil organic carbon forstakeholder decision-making ndash a case study from Kenya

Tor-Gunnar Varinggen Leigh Ann Winowiecki Constance Neely Sabrina Chesterman and Mieke BourneWorld Agroforestry Centre (ICRAF) PO Box 30677-00100 Nairobi Kenya

Correspondence Tor-Gunnar Varinggen (tvagencgiarorg)

Received 15 December 2017 ndash Discussion started 17 January 2018Revised 11 July 2018 ndash Accepted 2 October 2018 ndash Published 7 November 2018

Abstract Land degradation impacts the health and livelihoods of about 15 billion people worldwide Giventhat the state of the environment and food security are strongly interlinked in tropical landscapes the increasingneed for land for food production urbanization and other uses poses several threats to sustainability in the longterm This paper demonstrates the integration of land and soil health maps with socioeconomic datasets into anonline open-access platform called the Resilience Diagnostic and Decision Support Tool for Turkana Countyin Kenya using the Stakeholder Approach to Risk Informed and Evidence Based Decision Making (SHARED)methodology The paper highlights the utility of spatial assessments of soil organic carbon (SOC) for monitoringland degradation neutrality (LDN) compliance understanding the drivers of SOC dynamics and inclusion ofthese in stakeholder decision-making The main objectives of this paper were to (1) demonstrate the applicationof a systematic approach for land health assessments including spatial mapping of soil organic carbon (2) showan operational interdisciplinary framework for assessing ecosystem health and (3) showcase the application ofevidence-based tools for stakeholder engagement using the SHARED approach Through the approaches andtools presented the paper addresses the increasing need for more integrated approaches when assessing andmanaging ecosystem health to meet the targets of the 2030 Agenda including Sustainable Development Goal(SDG) 153 In addition to systematic and reliable biophysical and socioeconomic assessments stakeholderengagement with evidence is crucial to support such integrated approaches

1 Introduction

The 2030 Agenda including multilateral environmentalagreements and the Sustainable Development Goals (SDGs)has set the stage for greater appreciation and understandingof the complex nature and interactions among environmen-tal challenges facing society (United Nations General As-sembly 2015) SDG 15 calls for the protection restorationand promotion of sustainable use of terrestrial ecosystemsand sustainably managed forests combating desertificationhalting and reversing land degradation and halting biodiver-sity loss At their core initiatives such as land degradationneutrality (LDN) and SDG 15 emphasize healthy ecosystemfunction and resilient landscapes which underpin healthyeconomies and societal well-being Because these issues areintrinsically inter-related decisions around soil and land can-not be taken in isolation Achieving the associated targets

requires a robust evidence base for measuring and monitor-ing land health indicators and associated land managementpractices coupled with local- and policy-level awareness ofthe importance of land health in supporting multiple sectorsAlso local capacity to implement monitor and assess theseindicators is needed including mechanisms for integratedcross-sectoral coordination and inclusive multi-stakeholdercollaboration to achieve impact

Many assessments of land health suffer from (i) disagree-ments about the definition of land degradationland health(ii) an abundance of indicators (Heink and Kowarik 2010)that are often not feasible to measure and hence opera-tionalize and (iii) a lack of rigorous science-based indicators(Niemeijer and de Groot 2008) and analytical frameworksComparable indicators are critical when assessing environ-mental conditions and progress made towards the mitigation

Published by Copernicus Publications on behalf of the European Geosciences Union

260 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

or avoidance of land degradation but they are also importantto effectively communicate information both to stakeholderssuch as farmers or advisory services and to policy makersIn addition to being comparable indicators for assessmentand monitoring of land health should be (a) science-based(b) readily measurable (quantifiable) (c) rapid (d) based onfield assessment across multiple scales (plot field landscaperegion) and (e) representative of the complex processes ofland degradation in landscapes (Dale and Beyeler 2001) Themethods tools and approaches presented in this paper offerrealistic and cost-effective options both for baseline assess-ments of LDN indicators and for the monitoring of progresstowards LDN targets Several assessments of methodologiesfor LDN compliance (Aynekulu et al 2017 Caspari et al2015 Stavi and Lal 2015) also highlight the need for con-sistent spatially explicit data over time

The Land Degradation Surveillance Framework (LDSF)is an example of a method that has been applied in a num-ber of projects in the global tropics to provide more rigorousscience-based assessments of land degradation risk and sta-tus as well as soil health (Varinggen et al 2013a) An importantindicator of soil health is soil organic carbon (SOC) and fielddatasets collected using the LDSF have been used to providespatially explicit assessments of SOC and other indicators ofland degradation processes such as soil erosion in additionto other soil functional properties (Varinggen et al 2012 2016Winowiecki et al 2016a b) These assessments can be usedto inform spatially explicit land and soil health monitoringsystems which are critical in order for countries to avoidland degradation or to restore ecosystems that are alreadydegraded Combining systematic data collection efforts withrigorous analytical frameworks evidence-based approachescan be used to engage stakeholders with interactive onlineuser-friendly platforms to inform national and internationalpolicy makers

Within this context the engagement of stakeholders be-comes an important element to prioritize investment strate-gies for accelerating the achievement of the SDGs TheStakeholder Approach to Risk Informed and Evidence BasedDecision Making (SHARED)1 emerged in response to thisneed and was developed around a number of key factorssteps and principles including (i) advancing a holistic orsystems view to raise awareness on the integrated natureof environmental social cultural and economic dimensionsand causal relationships (ii) establishing a clear understand-ing of the influencing factors of human and group decision-making including stakeholder analysis (iii) facilitating dif-ferent government sectors and multi-stakeholder platforms ofdiverse societal sectors (iv) collectively articulating mutu-ally agreed desired sustainable development outcomes andindicators building upon fundamental ecosystem servicesand nested within national and global goals (v) generating

1httpwwwworldagroforestryorgshared (last access 21 Au-gust 2018)

evidence and experience and tailoring tools in a readily con-sumable way for problem solving and options identification(vi) testing options based on collectively defined criteria in-cluding risks and potential synergies and (vii) designing op-tion implementation with monitoring and evaluation and co-learning feedback into the process

This paper demonstrates the integration of land and soilhealth maps using spatial assessments of SOC as an exam-ple with socioeconomic datasets into an online tool the Re-silience Diagnostic and Decision Support Tool2 (RDDST) inTurkana County Kenya The main objectives of the effortsdiscussed in this paper were to (1) demonstrate the applica-tion of a systematic approach for land health assessmentsincluding spatial mapping of soil organic carbon (2) demon-strate the operationalization of interdisciplinary frameworkfor assessing ecosystem health and (3) showcase the applica-tion of evidence-based tools for stakeholder engagement us-ing the SHARED approach The approaches presented canalso be applied as part of LDN efforts to assist countriesin more effectively restoring degraded lands and increas-ing agricultural productivity including opportunities for sus-tainable development and biodiversity conservation (IUCN2015) A recent assessment of land degradation in Kenyaestimates that the annual cost of land degradation is at ap-proximately USD 15 billion which represents about 5 ofthe countryrsquos GDP (Munoz 2016) Kenya has endorsed theSDG 153 targets and discussions on baseline assessmentsand monitoring are underway Given the large extent of dry-lands in Kenya their management will be critical to achiev-ing these targets

2 Methodology

This paper utilizes spatial assessments of SOC that wereconducted using data from a network of LDSF sites in theglobal tropics The LDSF was designed for practical andcost-effective soil and ecosystem health surveillance includ-ing for the mapping of SOC (Varinggen et al 2013a 2016Winowiecki et al 2016a b) The framework is also designedto monitor changes over time and provides opportunitiesfor targeting improved soil management and land restora-tion activities (Lohbeck et al 2018) Specifically the LDSFsystematically assesses several ecological metrics simulta-neously at four different spatial scales (100 m2 1000 m21 km2 and 100 km2) using a spatially stratified hierarchi-cal sampling design (Varinggen et al 2013b) The LDSF alsoapplies the latest soil infrared (IR) spectroscopy technolo-gies in analysis of SOC and other soil properties which arecost-effective and hence allow for the scaling of soil mea-surements The spatial assessments of SOC included in thisstudy were developed based on an archive of 10 000 georef-erenced LDSF plots with soil samples that were analyzed

2httplandscapeportalorgsharedApp (last access 25 August2018)

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

260 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

or avoidance of land degradation but they are also importantto effectively communicate information both to stakeholderssuch as farmers or advisory services and to policy makersIn addition to being comparable indicators for assessmentand monitoring of land health should be (a) science-based(b) readily measurable (quantifiable) (c) rapid (d) based onfield assessment across multiple scales (plot field landscaperegion) and (e) representative of the complex processes ofland degradation in landscapes (Dale and Beyeler 2001) Themethods tools and approaches presented in this paper offerrealistic and cost-effective options both for baseline assess-ments of LDN indicators and for the monitoring of progresstowards LDN targets Several assessments of methodologiesfor LDN compliance (Aynekulu et al 2017 Caspari et al2015 Stavi and Lal 2015) also highlight the need for con-sistent spatially explicit data over time

The Land Degradation Surveillance Framework (LDSF)is an example of a method that has been applied in a num-ber of projects in the global tropics to provide more rigorousscience-based assessments of land degradation risk and sta-tus as well as soil health (Varinggen et al 2013a) An importantindicator of soil health is soil organic carbon (SOC) and fielddatasets collected using the LDSF have been used to providespatially explicit assessments of SOC and other indicators ofland degradation processes such as soil erosion in additionto other soil functional properties (Varinggen et al 2012 2016Winowiecki et al 2016a b) These assessments can be usedto inform spatially explicit land and soil health monitoringsystems which are critical in order for countries to avoidland degradation or to restore ecosystems that are alreadydegraded Combining systematic data collection efforts withrigorous analytical frameworks evidence-based approachescan be used to engage stakeholders with interactive onlineuser-friendly platforms to inform national and internationalpolicy makers

Within this context the engagement of stakeholders be-comes an important element to prioritize investment strate-gies for accelerating the achievement of the SDGs TheStakeholder Approach to Risk Informed and Evidence BasedDecision Making (SHARED)1 emerged in response to thisneed and was developed around a number of key factorssteps and principles including (i) advancing a holistic orsystems view to raise awareness on the integrated natureof environmental social cultural and economic dimensionsand causal relationships (ii) establishing a clear understand-ing of the influencing factors of human and group decision-making including stakeholder analysis (iii) facilitating dif-ferent government sectors and multi-stakeholder platforms ofdiverse societal sectors (iv) collectively articulating mutu-ally agreed desired sustainable development outcomes andindicators building upon fundamental ecosystem servicesand nested within national and global goals (v) generating

1httpwwwworldagroforestryorgshared (last access 21 Au-gust 2018)

evidence and experience and tailoring tools in a readily con-sumable way for problem solving and options identification(vi) testing options based on collectively defined criteria in-cluding risks and potential synergies and (vii) designing op-tion implementation with monitoring and evaluation and co-learning feedback into the process

This paper demonstrates the integration of land and soilhealth maps using spatial assessments of SOC as an exam-ple with socioeconomic datasets into an online tool the Re-silience Diagnostic and Decision Support Tool2 (RDDST) inTurkana County Kenya The main objectives of the effortsdiscussed in this paper were to (1) demonstrate the applica-tion of a systematic approach for land health assessmentsincluding spatial mapping of soil organic carbon (2) demon-strate the operationalization of interdisciplinary frameworkfor assessing ecosystem health and (3) showcase the applica-tion of evidence-based tools for stakeholder engagement us-ing the SHARED approach The approaches presented canalso be applied as part of LDN efforts to assist countriesin more effectively restoring degraded lands and increas-ing agricultural productivity including opportunities for sus-tainable development and biodiversity conservation (IUCN2015) A recent assessment of land degradation in Kenyaestimates that the annual cost of land degradation is at ap-proximately USD 15 billion which represents about 5 ofthe countryrsquos GDP (Munoz 2016) Kenya has endorsed theSDG 153 targets and discussions on baseline assessmentsand monitoring are underway Given the large extent of dry-lands in Kenya their management will be critical to achiev-ing these targets

2 Methodology

This paper utilizes spatial assessments of SOC that wereconducted using data from a network of LDSF sites in theglobal tropics The LDSF was designed for practical andcost-effective soil and ecosystem health surveillance includ-ing for the mapping of SOC (Varinggen et al 2013a 2016Winowiecki et al 2016a b) The framework is also designedto monitor changes over time and provides opportunitiesfor targeting improved soil management and land restora-tion activities (Lohbeck et al 2018) Specifically the LDSFsystematically assesses several ecological metrics simulta-neously at four different spatial scales (100 m2 1000 m21 km2 and 100 km2) using a spatially stratified hierarchi-cal sampling design (Varinggen et al 2013b) The LDSF alsoapplies the latest soil infrared (IR) spectroscopy technolo-gies in analysis of SOC and other soil properties which arecost-effective and hence allow for the scaling of soil mea-surements The spatial assessments of SOC included in thisstudy were developed based on an archive of 10 000 georef-erenced LDSF plots with soil samples that were analyzed

2httplandscapeportalorgsharedApp (last access 25 August2018)

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 261

for SOC and mid-infrared analysis at the ICRAF Soil andPlant Diagnostics Lab in Nairobi Kenya and used to con-duct spatial predictions of SOC concentrations and stocks us-ing reflectance data from the Moderate Resolution ImagingSpectroradiometer (MODIS) satellite sensor (Varinggen et al2016) In brief annual MODIS reflectance image compos-ites were created based on the date of maximum fractionalvegetation cover in each MODIS pixel and these reflectancedata were used to train models using the Random Forest al-gorithm (Breiman 2001) to predict soil properties such asSOC and pH as well as land degradation processes such assoil erosion The modeling of soil and land health variableswas conducted in R Statistics (R Core Team 2018) using therandomForest (Liaw and Wiener 2002) R library

An interactive dashboard was developed for TurkanaCounty which is located in northwestern Kenya borderingUganda Sudan and Ethiopia Turkana County is part of theArid and Semi-Arid Lands (ASALs) of Kenya with an an-nual average rainfall of about 250 mm and a population of al-most 1 million people the majority of which are pastoralistsWe used the Shiny web framework for R statistics (Changet al 2017 R Core Team 2018) to develop an interactivedashboard integrating existing and new data and providingrobust data management and graphical tools to allow usersto interact with these data in a meaningful way The RDDSTdashboard is hosted on the ICRAF Landscape Portal3 whichprovides a platform and tools for spatial data storage and vi-sualization

The SHARED approach was applied to foster the innova-tion required to shape and embed land health assessmentsinto inclusive negotiation and decision-making processesthrough a co-design process with stakeholders in TurkanaCounty By utilizing a comprehensive framework tailoredto specific decision needs coupled with science-based evi-dence processes evidence and tools were brought togetherto shift the decision paradigm towards more inclusive inter-sectoral and interinstitutional integration to tackle complexdecisions and achieve desired outcomes The approach em-phasizes the importance of rigorous indicator frameworksthat are simple and readily measurable but that reflect thecomplexity of both ecological systems and decision-makingprocesses

Through robust data management and advanced data visu-alization the RDDST was co-developed with stakeholders inTurkana County to help facilitate communication of data andanalysis between scientists and stakeholders This allows forthe interrogation of evidence and an increased rate of discov-ery and it helps contextualize the data used Third-party datasources for the RDDST included the Armed Conflict Loca-tion amp Event Data Project (ACLED) for security data Kenyagovernment statistics for education and health data and datafrom the Hunger Safety Net Programme (HSNP) for infor-

3httplandscapeportalorgsharedApp (last access 25 Septem-ber 2018)

Figure 1 The distribution of SOC stocks to 30 cm depth in KenyaThe vertical dashed line shows the mean (42 Mg haminus1) for the coun-try

mation on livestock populations and energy use within thecounty

The targeted facilitation and co-design approach employedensures local ownership and cohesive communication acrossmultiple institutions including different political levels andknowledge systems to develop capacity and the evidencebase as a continuously linked process At the national levelthe SHARED approach has been used in collaboration withKenyarsquos Ministry of Environment and Natural Resourcesand the Ministry of Agriculture Livestock and Fisheries tosynthesize the evidence of 44 integrated cropndashlivestockndashtreesystem projects to develop recommendations for climate-resilient approaches (Chesterman and Neely 2015) and todirectly inform the drafting of Kenyarsquos National ClimateChange Policy Framework (Neely 2014)

3 Results

Spatial estimates of SOC stocks for Kenya show that Kenyansoils store approximately 24 Gt carbon (C) in the upper30 cm of the soil profile (Minasny et al 2017) AverageSOC stocks for Kenya were estimated to be 42 Mg C haminus1

(Fig 1) for 0 to 30 cm depth The lowest estimated C stockswere found in arid and semi-arid regions of the country(lt20 Mg C haminus1 on average) such as Turkana County Theseestimates are consistent with those reported by Batjes andBatjes (2004) for arid and semi-arid parts of the countryHigher SOC stocks were found in subhumid and humid ar-eas including the central highlands and western parts of thecountry (Fig 1) The highest SOC stocks are found in forestsystems such as around Mt Kenya (gt100 Mg C haminus1) theAberdares the Mau Forest Complex and Kakamega ForestThese forest systems while making up a small proportion of

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

262 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

Figure 2 Soil organic carbon (SOC) map of Kenya at 500 m res-olution (for 2012) Turkana County is located in the northwesterncorner of the country and is outlined on the map

the countryrsquos area are critical carbon pools and also repre-sent important water towers that millions of Kenyans rely onfor their water supply (Mogaka et al 2005) Also wetlandecosystems such as inland riverine and palustrine areas arecritical SOC pools in Kenya and in much of Eastern AfricaIn the drylands such wetland systems are particularly crit-ical for SOC storage and land health frequently reachingbetween 80 and 100 Mg C haminus1 at 0 to 30 cm depth Otherwetland systems many of them under threat such as aroundthe Rift Valley lakes and lacustrine wetlands along Kenyarsquoscoast are also examples of ecosystems that are critical forC storage (Minasny et al 2017 Saunders et al 2007) andother ecosystem services (Zedler and Kercher 2005) in thecountry

Soil organic carbon (SOC) concentrations were mappedfor Africa at 500 m spatial resolution (Varinggen et al 2016)and clipped for Kenya (Fig 2) for use in stakeholder engage-ment processes within Turkana County (see county outlinedin Fig 2) Large parts of the drylands have SOC concen-trations that are lower than 15 g C kgminus1 which is generallyconsidered very low as shown in Fig 3 Given the emphasison management of SOC as a critical resource for both agri-culture and rangeland management in Turkana County SOCconcentrations were applied in the RDDST rather than esti-mates of SOC stocks

Concentrations of SOC follow a similar pattern as the SOCstocks as shown in Figs 2 and 3 with very low SOC in partsof the drier northern and northeastern parts of the countryHowever it is worth noting that SOC is relatively heteroge-

Figure 3 Histograms showing the distribution of SOC in TurkanaCounty (green) versus the rest of Kenya (red) The vertical dashedline represents a SOC concentration of 15 g kgminus1

nous within these areas and that there are pockets of highSOC in parts of the drylands such as the Matthews Rangeand mountains such as Ndoto Marsabit and Kulal as wellas the Loima Hills in Turkana County These areas representcritical SOC pools are important resources for pastoralistsin the drylands for dry-season grazing (Oba et al 2000) andare hotspots for biodiversity in the region

In Turkana County maps of SOC concentrations were in-cluded in the RDDST tool to assess resilience within thecounty and support decision-making processes (Fig 4) in-tegrating data from across numerous sectors from securityand education to human health and land health Through aserious of workshops facilitated using the SHARED pro-cess the RDDST was co-developed with representativesfrom the Turkana County Government and other stakehold-ers such as representatives from UN organizations and non-governmental organizations (NGOs) in the county Further-more the outcomes of these workshops and the RDDSTwere applied in the development of the Turkana County In-tegrated Development Plan (CIDP) for the period 2018 to2022 The land health part of the RDDST was used to inter-actively visualize and explore SOC (Fig 5) along with otherkey land health indicators across the county including soilerosion prevalence root-depth restrictions and soil pH us-ing spatial estimates from Varinggen et al (2016) In additiona module for exploring vegetation cover status and trendswas applied for assessing vegetation performance and trendsThese tools are being used to demonstrate the critical role ofthe underpinning natural resource base and the importance oftaking this into account in decision-making across multiplesectors Visualization of local land health parameters in con-junction with data from other sectors shifted the discourse inthe county government and resulted in the identification of

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 263

Figure 4 The front page of the Turkana Resilience Diagnostic and Decision Support Tool

integrated flagships for multiple sectors working together ondevelopment outcomes with a base in land restoration

4 Discussion

Consistently developed indicator frameworks such as theLDSF combined with systematic field and lab measure-ment protocols allow for spatial assessments of SOC andother land health indicators at multiple spatial scales andwith unprecedented accuracy and spatial coverage Such spa-tial assessments can be applied at continental and regionalscales (Varinggen et al 2016) as well as at country and lo-cal scales (Figs 2 and 3) with results integrated directlyinto decision-making processes at a local level as demon-strated for Turkana County (Fig 5) In Kenyan drylandswhich make up more than 80 of the area of the countryand where the primary land use is livestock production SOCconcentrations and stocks are low due to a combination of

climatic conditions and widespread land degradation Landdegradation in these drylands is the result of several factorsincluding substantial migration into these areas which hasresulted in a large growth in their population starting as earlyas over 30 years ago (Bernard 1985) and soils that are alka-line and prone to soil erosion This along with other factorshas in turn resulted in overgrazing or inappropriate agricul-tural practices that have again resulted in accelerated landdegradation As a result counties in the drylands of Kenyaface serious challenges in terms of restoring degraded landsincluding implementing policies and incentives that can helptarget restoration efforts both in terms of space and contextBy applying spatial assessments of land health indicatorsSOC being a key indicator in this regard areas that are de-graded can be identified for further assessments of the driversand processes that are leading to degradation while at thesame time making more realistic assessments of the types ofrestoration efforts that are likely to be effective given condi-tions in a certain area or location

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

264 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

Figure 5 The SOC assessment part of the Turkana Resilience Diagnostic and Decision Support Tool which is part of the land health sectionof the tool

There is great value in using the soil health maps depict-ing SOC and other attributes for awareness raising and toinform decisions of stakeholders and decision makers Build-ing these data into the RDDST and displaying them in con-junction with data and statistics from other sectors includ-ing human health education pastoral economies environ-ment agriculture trade infrastructure and energy supportedthe facilitation of a newfound attention to the socioecolog-ical system and associated inter-relationships and an under-standing of causal dynamics among the sectors The Ministryof Economic Planning working with all of the governmentsectors and several development actors in Turkana County

chose to use the SHARED methodology and the RDDST toguide the new Turkana County Integrated Development Plan(CIDP) a 5-year program (2017ndash2022) Based on the knowl-edge of the prevalence and location of land degradation inthe county there were multiple county-integrated flagshipsaimed at restoring land health while meeting the priorities ofsocial and economic sectors Further the county governmentchose to cancel an earlier planned infrastructural effort thatwould have exposed relatively healthy land to degradationPresenting readily accessible soil health maps initiated dis-cussion on the fundamental importance of land restoration toachieving development outcomes and led to a commitment to

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 265

bring land health management into integrated planning pro-cesses within the county

5 Conclusions

Recent developments in the use of earth observation data tomodel and map SOC concentrations and stocks have resultedin the availability of spatial estimates with high accuracy andmoderate to fine spatial resolution Decision support tools in-crease the utility of spatial assessments of SOC along withother indicators of soil and land health by providing userswith both interactive dashboards that allow them to map andinteract with data and analytical tools for land health diag-nostics and the targeting of interventions Through the ap-plication of open-source platforms and tools the use of ev-idence in land health management can also be effectivelymainstreamed In the context of the SHARED approach thisprocess is enhanced through structured stakeholder engage-ment co-learning and co-design of tools These approachesand tools have the potential to greatly enhance the uptake ofmanagement interventions to increase SOC and avoid or re-verse land degradation hence also contributing towards theachievement of LDN and SDG targets

Data availability Maps were produced using a dataset which isavailable in Harvard Dataverse (Varinggen et al 2013c) The SOC mapof Kenya will be available on the ICRAF Landscape Portal to allregistered users at httplandscapeportalorg registration is free

Supplement The supplement related to this article is availableonline at httpsdoiorg105194soil-4-259-2018-supplement

Author contributions TGV led the development of the data ana-lytics and dashboards as well the predictive mapping of SOC TGValso led the writing of the manuscript LAW contributed to the con-ceptual framework of the paper data analytics and writing and edit-ing of the paper CN led the SHARED facilitation workshops co-designed the dashboards and contributed to the framing and writingof the manuscript SC co-led the SHARED facilitation around thedashboards and the graphical design and contributed to the theoreti-cal basis of SHARED MB co-led the SHARED workshops and theconcepts employed in SHARED and contributed to the writing ofthe manuscript

Competing interests The authors declare that they have no con-flict of interest

Special issue statement This article is part of the special issueldquoRegional perspectives and challenges of soil organic carbon man-agement and monitoring ndash a special issue from the Global Sympo-

sium on Soil Organic Carbon 2017rdquo It is a result of the Global Sym-posium on Soil Organic Carbon Rome Italy 21ndash23 March 2017

Acknowledgements This work was supported in part by theInternational Fund for Agricultural Development (IFAD) grantnumbers 2000000520 2000000976 and 2000001302 US Agencyfor International Development Resilience Program UNICEFas well as the CGIAR Research Program on Forests Trees andAgroforestry (FTA) We also acknowledge partners from theTurkana County Government in Kenya

Edited by Viridiana AlcaacutentaraReviewed by two anonymous referees

References

Aynekulu E Lohbeck M Nijbroek R Ordoacutentildeez J Turner KVaringgen T-G and Winowiecki L A Review of methodolo-gies for land degradation neutrality baselines Sub-national casestudies from Costa Rica and Namibia 58 available at httphdlhandlenet1056880563 2017

Batjes N and Batjes N Soil carbon stocks and pro-jected changes according to land use and management acase study for Kenya Soil Use Manage 20 350ndash356httpsdoiorg101079SUM2004269 2004

Bernard F E Planning and Environmental Risk in Kenyan Dry-lands Geogr Rev 75 58 httpsdoiorg102307214578 1985

Breiman L Random forests Mach Learn 45 35httpsdoiorg101023A1010933404324 2001

Caspari T van Lynden G and Bai Z Land Degradation Neutral-ity An Evaluation of Methods ISRIC-World Soil InformationWageningen 2015

Chang W Cheng J Allaire J J Xie Y and McPherson Jshiny Web Application Framework for R available at httpscranr-projectorgpackage=shiny (last access 20 August 2018)2017

Chesterman S and Neely C Evidence into decision makingfor resilience planning in Turkana County Stakeholder Ap-proach for Risk Informed and Evidence Based Decision Mak-ing (SHARED) World Agroforestry Centre (ICRAF) WorkingPaper 13 pp 2015

Dale V H and Beyeler S C Challenges in the develop-ment and use of ecological indicators Ecol Indic 1 3ndash10httpsdoiorg101016S1470-160X(01)00003-6 2001

Heink U and Kowarik I What are indicators On the definitionof indicators in ecology and environmental planning Ecol In-dic 10 584ndash593 httpsdoiorg101016jecolind2009090092010

IUCN Land Degradation Neutrality Implications and opportunitiesfor conservation Nairobi Kenya 2015

Liaw A and Wiener M Classification and Regres-sion by randomForest R news available at ftp1312529779TransferTregWFRE_ArticlesLiaw_02_ClassificationandregressionbyrandomForestpdf (last access 25August 2018) 2002

Lohbeck M Winowiecki L Aynekulu E Okia C and Varing-gen T-G Trait-based approaches for guiding the restoration of

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 263

Figure 4 The front page of the Turkana Resilience Diagnostic and Decision Support Tool

integrated flagships for multiple sectors working together ondevelopment outcomes with a base in land restoration

4 Discussion

Consistently developed indicator frameworks such as theLDSF combined with systematic field and lab measure-ment protocols allow for spatial assessments of SOC andother land health indicators at multiple spatial scales andwith unprecedented accuracy and spatial coverage Such spa-tial assessments can be applied at continental and regionalscales (Varinggen et al 2016) as well as at country and lo-cal scales (Figs 2 and 3) with results integrated directlyinto decision-making processes at a local level as demon-strated for Turkana County (Fig 5) In Kenyan drylandswhich make up more than 80 of the area of the countryand where the primary land use is livestock production SOCconcentrations and stocks are low due to a combination of

climatic conditions and widespread land degradation Landdegradation in these drylands is the result of several factorsincluding substantial migration into these areas which hasresulted in a large growth in their population starting as earlyas over 30 years ago (Bernard 1985) and soils that are alka-line and prone to soil erosion This along with other factorshas in turn resulted in overgrazing or inappropriate agricul-tural practices that have again resulted in accelerated landdegradation As a result counties in the drylands of Kenyaface serious challenges in terms of restoring degraded landsincluding implementing policies and incentives that can helptarget restoration efforts both in terms of space and contextBy applying spatial assessments of land health indicatorsSOC being a key indicator in this regard areas that are de-graded can be identified for further assessments of the driversand processes that are leading to degradation while at thesame time making more realistic assessments of the types ofrestoration efforts that are likely to be effective given condi-tions in a certain area or location

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

264 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

Figure 5 The SOC assessment part of the Turkana Resilience Diagnostic and Decision Support Tool which is part of the land health sectionof the tool

There is great value in using the soil health maps depict-ing SOC and other attributes for awareness raising and toinform decisions of stakeholders and decision makers Build-ing these data into the RDDST and displaying them in con-junction with data and statistics from other sectors includ-ing human health education pastoral economies environ-ment agriculture trade infrastructure and energy supportedthe facilitation of a newfound attention to the socioecolog-ical system and associated inter-relationships and an under-standing of causal dynamics among the sectors The Ministryof Economic Planning working with all of the governmentsectors and several development actors in Turkana County

chose to use the SHARED methodology and the RDDST toguide the new Turkana County Integrated Development Plan(CIDP) a 5-year program (2017ndash2022) Based on the knowl-edge of the prevalence and location of land degradation inthe county there were multiple county-integrated flagshipsaimed at restoring land health while meeting the priorities ofsocial and economic sectors Further the county governmentchose to cancel an earlier planned infrastructural effort thatwould have exposed relatively healthy land to degradationPresenting readily accessible soil health maps initiated dis-cussion on the fundamental importance of land restoration toachieving development outcomes and led to a commitment to

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 265

bring land health management into integrated planning pro-cesses within the county

5 Conclusions

Recent developments in the use of earth observation data tomodel and map SOC concentrations and stocks have resultedin the availability of spatial estimates with high accuracy andmoderate to fine spatial resolution Decision support tools in-crease the utility of spatial assessments of SOC along withother indicators of soil and land health by providing userswith both interactive dashboards that allow them to map andinteract with data and analytical tools for land health diag-nostics and the targeting of interventions Through the ap-plication of open-source platforms and tools the use of ev-idence in land health management can also be effectivelymainstreamed In the context of the SHARED approach thisprocess is enhanced through structured stakeholder engage-ment co-learning and co-design of tools These approachesand tools have the potential to greatly enhance the uptake ofmanagement interventions to increase SOC and avoid or re-verse land degradation hence also contributing towards theachievement of LDN and SDG targets

Data availability Maps were produced using a dataset which isavailable in Harvard Dataverse (Varinggen et al 2013c) The SOC mapof Kenya will be available on the ICRAF Landscape Portal to allregistered users at httplandscapeportalorg registration is free

Supplement The supplement related to this article is availableonline at httpsdoiorg105194soil-4-259-2018-supplement

Author contributions TGV led the development of the data ana-lytics and dashboards as well the predictive mapping of SOC TGValso led the writing of the manuscript LAW contributed to the con-ceptual framework of the paper data analytics and writing and edit-ing of the paper CN led the SHARED facilitation workshops co-designed the dashboards and contributed to the framing and writingof the manuscript SC co-led the SHARED facilitation around thedashboards and the graphical design and contributed to the theoreti-cal basis of SHARED MB co-led the SHARED workshops and theconcepts employed in SHARED and contributed to the writing ofthe manuscript

Competing interests The authors declare that they have no con-flict of interest

Special issue statement This article is part of the special issueldquoRegional perspectives and challenges of soil organic carbon man-agement and monitoring ndash a special issue from the Global Sympo-

sium on Soil Organic Carbon 2017rdquo It is a result of the Global Sym-posium on Soil Organic Carbon Rome Italy 21ndash23 March 2017

Acknowledgements This work was supported in part by theInternational Fund for Agricultural Development (IFAD) grantnumbers 2000000520 2000000976 and 2000001302 US Agencyfor International Development Resilience Program UNICEFas well as the CGIAR Research Program on Forests Trees andAgroforestry (FTA) We also acknowledge partners from theTurkana County Government in Kenya

Edited by Viridiana AlcaacutentaraReviewed by two anonymous referees

References

Aynekulu E Lohbeck M Nijbroek R Ordoacutentildeez J Turner KVaringgen T-G and Winowiecki L A Review of methodolo-gies for land degradation neutrality baselines Sub-national casestudies from Costa Rica and Namibia 58 available at httphdlhandlenet1056880563 2017

Batjes N and Batjes N Soil carbon stocks and pro-jected changes according to land use and management acase study for Kenya Soil Use Manage 20 350ndash356httpsdoiorg101079SUM2004269 2004

Bernard F E Planning and Environmental Risk in Kenyan Dry-lands Geogr Rev 75 58 httpsdoiorg102307214578 1985

Breiman L Random forests Mach Learn 45 35httpsdoiorg101023A1010933404324 2001

Caspari T van Lynden G and Bai Z Land Degradation Neutral-ity An Evaluation of Methods ISRIC-World Soil InformationWageningen 2015

Chang W Cheng J Allaire J J Xie Y and McPherson Jshiny Web Application Framework for R available at httpscranr-projectorgpackage=shiny (last access 20 August 2018)2017

Chesterman S and Neely C Evidence into decision makingfor resilience planning in Turkana County Stakeholder Ap-proach for Risk Informed and Evidence Based Decision Mak-ing (SHARED) World Agroforestry Centre (ICRAF) WorkingPaper 13 pp 2015

Dale V H and Beyeler S C Challenges in the develop-ment and use of ecological indicators Ecol Indic 1 3ndash10httpsdoiorg101016S1470-160X(01)00003-6 2001

Heink U and Kowarik I What are indicators On the definitionof indicators in ecology and environmental planning Ecol In-dic 10 584ndash593 httpsdoiorg101016jecolind2009090092010

IUCN Land Degradation Neutrality Implications and opportunitiesfor conservation Nairobi Kenya 2015

Liaw A and Wiener M Classification and Regres-sion by randomForest R news available at ftp1312529779TransferTregWFRE_ArticlesLiaw_02_ClassificationandregressionbyrandomForestpdf (last access 25August 2018) 2002

Lohbeck M Winowiecki L Aynekulu E Okia C and Varing-gen T-G Trait-based approaches for guiding the restoration of

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

264 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

Figure 5 The SOC assessment part of the Turkana Resilience Diagnostic and Decision Support Tool which is part of the land health sectionof the tool

There is great value in using the soil health maps depict-ing SOC and other attributes for awareness raising and toinform decisions of stakeholders and decision makers Build-ing these data into the RDDST and displaying them in con-junction with data and statistics from other sectors includ-ing human health education pastoral economies environ-ment agriculture trade infrastructure and energy supportedthe facilitation of a newfound attention to the socioecolog-ical system and associated inter-relationships and an under-standing of causal dynamics among the sectors The Ministryof Economic Planning working with all of the governmentsectors and several development actors in Turkana County

chose to use the SHARED methodology and the RDDST toguide the new Turkana County Integrated Development Plan(CIDP) a 5-year program (2017ndash2022) Based on the knowl-edge of the prevalence and location of land degradation inthe county there were multiple county-integrated flagshipsaimed at restoring land health while meeting the priorities ofsocial and economic sectors Further the county governmentchose to cancel an earlier planned infrastructural effort thatwould have exposed relatively healthy land to degradationPresenting readily accessible soil health maps initiated dis-cussion on the fundamental importance of land restoration toachieving development outcomes and led to a commitment to

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 265

bring land health management into integrated planning pro-cesses within the county

5 Conclusions

Recent developments in the use of earth observation data tomodel and map SOC concentrations and stocks have resultedin the availability of spatial estimates with high accuracy andmoderate to fine spatial resolution Decision support tools in-crease the utility of spatial assessments of SOC along withother indicators of soil and land health by providing userswith both interactive dashboards that allow them to map andinteract with data and analytical tools for land health diag-nostics and the targeting of interventions Through the ap-plication of open-source platforms and tools the use of ev-idence in land health management can also be effectivelymainstreamed In the context of the SHARED approach thisprocess is enhanced through structured stakeholder engage-ment co-learning and co-design of tools These approachesand tools have the potential to greatly enhance the uptake ofmanagement interventions to increase SOC and avoid or re-verse land degradation hence also contributing towards theachievement of LDN and SDG targets

Data availability Maps were produced using a dataset which isavailable in Harvard Dataverse (Varinggen et al 2013c) The SOC mapof Kenya will be available on the ICRAF Landscape Portal to allregistered users at httplandscapeportalorg registration is free

Supplement The supplement related to this article is availableonline at httpsdoiorg105194soil-4-259-2018-supplement

Author contributions TGV led the development of the data ana-lytics and dashboards as well the predictive mapping of SOC TGValso led the writing of the manuscript LAW contributed to the con-ceptual framework of the paper data analytics and writing and edit-ing of the paper CN led the SHARED facilitation workshops co-designed the dashboards and contributed to the framing and writingof the manuscript SC co-led the SHARED facilitation around thedashboards and the graphical design and contributed to the theoreti-cal basis of SHARED MB co-led the SHARED workshops and theconcepts employed in SHARED and contributed to the writing ofthe manuscript

Competing interests The authors declare that they have no con-flict of interest

Special issue statement This article is part of the special issueldquoRegional perspectives and challenges of soil organic carbon man-agement and monitoring ndash a special issue from the Global Sympo-

sium on Soil Organic Carbon 2017rdquo It is a result of the Global Sym-posium on Soil Organic Carbon Rome Italy 21ndash23 March 2017

Acknowledgements This work was supported in part by theInternational Fund for Agricultural Development (IFAD) grantnumbers 2000000520 2000000976 and 2000001302 US Agencyfor International Development Resilience Program UNICEFas well as the CGIAR Research Program on Forests Trees andAgroforestry (FTA) We also acknowledge partners from theTurkana County Government in Kenya

Edited by Viridiana AlcaacutentaraReviewed by two anonymous referees

References

Aynekulu E Lohbeck M Nijbroek R Ordoacutentildeez J Turner KVaringgen T-G and Winowiecki L A Review of methodolo-gies for land degradation neutrality baselines Sub-national casestudies from Costa Rica and Namibia 58 available at httphdlhandlenet1056880563 2017

Batjes N and Batjes N Soil carbon stocks and pro-jected changes according to land use and management acase study for Kenya Soil Use Manage 20 350ndash356httpsdoiorg101079SUM2004269 2004

Bernard F E Planning and Environmental Risk in Kenyan Dry-lands Geogr Rev 75 58 httpsdoiorg102307214578 1985

Breiman L Random forests Mach Learn 45 35httpsdoiorg101023A1010933404324 2001

Caspari T van Lynden G and Bai Z Land Degradation Neutral-ity An Evaluation of Methods ISRIC-World Soil InformationWageningen 2015

Chang W Cheng J Allaire J J Xie Y and McPherson Jshiny Web Application Framework for R available at httpscranr-projectorgpackage=shiny (last access 20 August 2018)2017

Chesterman S and Neely C Evidence into decision makingfor resilience planning in Turkana County Stakeholder Ap-proach for Risk Informed and Evidence Based Decision Mak-ing (SHARED) World Agroforestry Centre (ICRAF) WorkingPaper 13 pp 2015

Dale V H and Beyeler S C Challenges in the develop-ment and use of ecological indicators Ecol Indic 1 3ndash10httpsdoiorg101016S1470-160X(01)00003-6 2001

Heink U and Kowarik I What are indicators On the definitionof indicators in ecology and environmental planning Ecol In-dic 10 584ndash593 httpsdoiorg101016jecolind2009090092010

IUCN Land Degradation Neutrality Implications and opportunitiesfor conservation Nairobi Kenya 2015

Liaw A and Wiener M Classification and Regres-sion by randomForest R news available at ftp1312529779TransferTregWFRE_ArticlesLiaw_02_ClassificationandregressionbyrandomForestpdf (last access 25August 2018) 2002

Lohbeck M Winowiecki L Aynekulu E Okia C and Varing-gen T-G Trait-based approaches for guiding the restoration of

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making 265

bring land health management into integrated planning pro-cesses within the county

5 Conclusions

Recent developments in the use of earth observation data tomodel and map SOC concentrations and stocks have resultedin the availability of spatial estimates with high accuracy andmoderate to fine spatial resolution Decision support tools in-crease the utility of spatial assessments of SOC along withother indicators of soil and land health by providing userswith both interactive dashboards that allow them to map andinteract with data and analytical tools for land health diag-nostics and the targeting of interventions Through the ap-plication of open-source platforms and tools the use of ev-idence in land health management can also be effectivelymainstreamed In the context of the SHARED approach thisprocess is enhanced through structured stakeholder engage-ment co-learning and co-design of tools These approachesand tools have the potential to greatly enhance the uptake ofmanagement interventions to increase SOC and avoid or re-verse land degradation hence also contributing towards theachievement of LDN and SDG targets

Data availability Maps were produced using a dataset which isavailable in Harvard Dataverse (Varinggen et al 2013c) The SOC mapof Kenya will be available on the ICRAF Landscape Portal to allregistered users at httplandscapeportalorg registration is free

Supplement The supplement related to this article is availableonline at httpsdoiorg105194soil-4-259-2018-supplement

Author contributions TGV led the development of the data ana-lytics and dashboards as well the predictive mapping of SOC TGValso led the writing of the manuscript LAW contributed to the con-ceptual framework of the paper data analytics and writing and edit-ing of the paper CN led the SHARED facilitation workshops co-designed the dashboards and contributed to the framing and writingof the manuscript SC co-led the SHARED facilitation around thedashboards and the graphical design and contributed to the theoreti-cal basis of SHARED MB co-led the SHARED workshops and theconcepts employed in SHARED and contributed to the writing ofthe manuscript

Competing interests The authors declare that they have no con-flict of interest

Special issue statement This article is part of the special issueldquoRegional perspectives and challenges of soil organic carbon man-agement and monitoring ndash a special issue from the Global Sympo-

sium on Soil Organic Carbon 2017rdquo It is a result of the Global Sym-posium on Soil Organic Carbon Rome Italy 21ndash23 March 2017

Acknowledgements This work was supported in part by theInternational Fund for Agricultural Development (IFAD) grantnumbers 2000000520 2000000976 and 2000001302 US Agencyfor International Development Resilience Program UNICEFas well as the CGIAR Research Program on Forests Trees andAgroforestry (FTA) We also acknowledge partners from theTurkana County Government in Kenya

Edited by Viridiana AlcaacutentaraReviewed by two anonymous referees

References

Aynekulu E Lohbeck M Nijbroek R Ordoacutentildeez J Turner KVaringgen T-G and Winowiecki L A Review of methodolo-gies for land degradation neutrality baselines Sub-national casestudies from Costa Rica and Namibia 58 available at httphdlhandlenet1056880563 2017

Batjes N and Batjes N Soil carbon stocks and pro-jected changes according to land use and management acase study for Kenya Soil Use Manage 20 350ndash356httpsdoiorg101079SUM2004269 2004

Bernard F E Planning and Environmental Risk in Kenyan Dry-lands Geogr Rev 75 58 httpsdoiorg102307214578 1985

Breiman L Random forests Mach Learn 45 35httpsdoiorg101023A1010933404324 2001

Caspari T van Lynden G and Bai Z Land Degradation Neutral-ity An Evaluation of Methods ISRIC-World Soil InformationWageningen 2015

Chang W Cheng J Allaire J J Xie Y and McPherson Jshiny Web Application Framework for R available at httpscranr-projectorgpackage=shiny (last access 20 August 2018)2017

Chesterman S and Neely C Evidence into decision makingfor resilience planning in Turkana County Stakeholder Ap-proach for Risk Informed and Evidence Based Decision Mak-ing (SHARED) World Agroforestry Centre (ICRAF) WorkingPaper 13 pp 2015

Dale V H and Beyeler S C Challenges in the develop-ment and use of ecological indicators Ecol Indic 1 3ndash10httpsdoiorg101016S1470-160X(01)00003-6 2001

Heink U and Kowarik I What are indicators On the definitionof indicators in ecology and environmental planning Ecol In-dic 10 584ndash593 httpsdoiorg101016jecolind2009090092010

IUCN Land Degradation Neutrality Implications and opportunitiesfor conservation Nairobi Kenya 2015

Liaw A and Wiener M Classification and Regres-sion by randomForest R news available at ftp1312529779TransferTregWFRE_ArticlesLiaw_02_ClassificationandregressionbyrandomForestpdf (last access 25August 2018) 2002

Lohbeck M Winowiecki L Aynekulu E Okia C and Varing-gen T-G Trait-based approaches for guiding the restoration of

wwwsoil-journalnet42592018 SOIL 4 259ndash266 2018

266 T-G Varinggen et al Spatial assessments of soil organic carbon for stakeholder decision-making

degraded agricultural landscapes in East Africa edited by MIsaac J Appl Ecol 55 59ndash68 httpsdoiorg1011111365-266413017 2018

Minasny B Malone B P McBratney A B Angers D AArrouays D Chambers A Chaplot V Chen Z-S ChengK Das B S Field D J Gimona A Hedley C BHong S Y Mandal B Marchant B P Martin M Mc-Conkey B G Mulder V L OrsquoRourke S Richer-de-ForgesA C Odeh I Padarian J Paustian K Pan G Pog-gio L Savin I Stolbovoy V Stockmann U SulaemanY Tsui C-C Varinggen T-G Wesemael B and van andWinowiecki L Soil carbon 4 per mille Geoderma 292 59ndash86httpsdoiorg101016jgeoderma201701002 2017

Mogaka H Gichere S Davis R and Hirji R Climate Variabil-ity and Water Resources Degradation in Kenya The World Bank2005

Munoz P Land Degradation Neutrality (LDN) in Kenya The Eco-nomic Case UNCCD Nairobi Kenya 2016

Neely C Transitioning towards Climate-Smart Agriculture inKenya Linking Research Practice and Policy FAO Brief 6 pphttpwwwfaoorg3a-i4259epdf (last access 20 August 2018)2014

Niemeijer D and de Groot R S A conceptual framework forselecting environmental indicator sets Ecol Indic 8 14ndash25httpsdoiorg101016jecolind200611012 2008

Oba G Stenseth N C and Lusigi W J New perspectives onsustainable grazing management in arid zones of sub-SaharanAfrica BioScience 50 35ndash51 httpsdoiorg1016410006-3568(2000)050[0035NPOSGM]23CO2 2000

R Core Team R A Language and Environment for Statistical Com-puting R Foundation for Statistical Computing Vienna Austriaavailable at httpswwwr-projectorg last access 25 August2018

Saunders M J Jones M B and Kansiime F Carbon and watercycles in tropical papyrus wetlands Wetl Ecol Manag 15 489ndash498 httpsdoiorg101007s11273-007-9051-9 2007

Stavi I and Lal R Achieving Zero Net Land Degradation Chal-lenges and opportunities J Arid Environ 112(PA) 44ndash51httpsdoiorg101016jjaridenv201401016 2015

United Nations General Assembly Transforming our world The2030 agenda for sustainable development 1 1ndash5 httpwwwunorggasearchview_docaspsymbol=ARES701ampLang=ELast(last access 10 July 2018) 2015

Varinggen T-G Davey F and Shepherd K D Land health surveil-lance Mapping soil carbon in Kenyan rangelands in Agro-forestry ndash the future of global land use edited by Nair P Rand Garrity D 455ndash462 Springer 2012

Varinggen T-G Winowiecki L A Abegaz A and HadguK M Landsat-based approaches for mapping of landdegradation prevalence and soil functional propertiesin Ethiopia Remote Sens Environ 134 266ndash275httpsdoiorg101016jrse201303006 2013a

Varinggen T-G Winowiecki L A Tamene Desta L and TondohJ E the Land Degradation Surveillance Framework (LDSF)ndash Field Guide 14 available at httplandscapeportalorgblog20150325the-land-degradation-surveillance-framework-ldsf(last access 15 August 2018) 2013b

Varinggen T-G Winowiecki L A Tondoh J E and Desta L TAfrica Soil Information Service (AfSIS) ndash Soil Health Map-ping httpshdlhandlenet1902119793 Harvard DataverseV2 2013c

Varinggen T-G Winowiecki L A Tondoh J E Desta L T andGumbricht T Mapping of soil properties and land degradationrisk in Africa using MODIS reflectance Geoderma 263 216ndash225 httpsdoiorg101016jgeoderma201506023 2016

Winowiecki L Varinggen T-G and Huising J Effects of landcover on ecosystem services in Tanzania A spatial as-sessment of soil organic carbon Geoderma 263 274ndash283httpsdoiorg101016jgeoderma201503010 2016a

Winowiecki L Varinggen T-G Massawe B Jelinski N A Lyam-chai C Sayula G and Msoka E Landscape-scale variabil-ity of soil health indicators effects of cultivation on soil or-ganic carbon in the Usambara Mountains of Tanzania Nutr CyclAgroecosyst 105 263ndash274 httpsdoiorg101007s10705-015-9750-1 2016b

Zedler J B and Kercher S WETLAND RE-SOURCES Status Trends Ecosystem Services andRestorability Annu Rev Env Resour 30 39ndash74httpsdoiorg101146annurevenergy300505041442482005

SOIL 4 259ndash266 2018 wwwsoil-journalnet42592018