3.3 Land-cover/ use change prediction
By using the VALIDATE method in the Terrset GIS software, agreement was measured between predicted maps and the classified map for 2016. The Standard Kappa Index, Quality Kappa Index (no location), and Location Kappa Index for 2016 model validation were 0.91, 0.92, and 0.93, respectively (table 4). The accuracy of the predicted map for 2016 was 88%.
The predicted LULCC map from 2016 to 2036 revealed multiple trajectories of land transformation, such as an increase of some classes often at the expense of rangeland type I. For instance, ‘residences’ should increase to 9000 (ha) at the expense of class ‘agriculture’ and ‘rangeland type I’ (figure 7, 8).
<Table 4>
<Figure 6>
<Figure 7>
<Figure 8>
Discussion
Land-cover/use change has various implications for the environment (De Rosa et al. , 2016). LULCC prediction models provide the possibility to assess the change of land cover and land use and supply important information for better land-use planning to managers and stakeholders (Ansari & Golabi, 2019; Herrmann & Osinski, 1999). Often, watersheds are considered as an important assessment unit for land planning. Therefore, we determined our study area based on the ecologic border of the Lenjan and Najafabad watersheds of the Zayandehrood River to prepare a suitable estimation of the spatial transition potential of the land. The study area is within the province administrative units and the results of the study can be used for the development of land management plans.
We reconstructed land cover/ use change with multiseasonal Landsat imagery for 1985, 1998 and 2016 for the watershed surrounding the Zayandehrood River, which is the main water resource of the study area and our study revealed multiple trajectories of land change. For instance, our study showed both agricultural expansion, but also abandonment (an increase of fallow and completely abandoned areas from 1985 to 2016). Several studies suggested that human factors, such as over-using underground water resources, and water transfer plans are the main reasons for water decrease including the Zayandehrood River and, consequently, resulting in abandonment of agriculture plots where irrigation and production costs on degraded lands are not compensated by the returns from farming (Azadi et al. , 2016; Fazli, A & Badi, Sh, 2017; Minaei et al. , 2018). Therefore, the current water management plans are not sustainable as they result in exhausting water resources and salinization of some plots as in other parts of Iran (Iranmehr et al. , 2015; Yaghobi et al. , 2019; Zeaieanet al. , 2005).
Our classification results also revealed the conversion of agricultural lands to residences and recreational and industrial areas in close proximity to residential areas, in drought-prone areas, and with deficient water resources. These findings were in line with earlier studies in other parts of Iran, which also found a contraction of agricultural lands in proximity to residential areas (Eisavi et al. , 2015; Mosammam et al. , 2017; Soffianian & Madanian, 2015). Such finds are worrisome because Iran also relies on food imports. Any additional contractions of croplands in Iran may pose a threat for food insecurity.
Our findings also showed that residential areas and even new industrial towns such as Majlesi were built at the expense of the rangelands. Contrary to the contraction of agricultural lands in the vicinity of residential areas, we also observed an increase in agricultural land at the expense of rangelands. Rangelands that were not converted to croplands, were often left at the marginal areas. Thus, any new cropland expansion at the expense of rangelands suggested cropland encroachments in agro-environmental frontiers with unstable yields and in areas prone to abandonment (Kraemer et al. , 2015; Yaghobi et al. , 2019). Rangelands in many parts of the world, including Iran, have a weak protective status. At the same time, rangelands host charismatic umbrella species and support the livelihoods in rural areas, including rangeland livestock herding (Behnke et al. , 2016; Dara et al. , 2020; Kerven et al. , 2006). Therefore, the permanent conversion of rangelands to industrial areas should be controlled, and zoning should be developed to avoid the unintended land conversions and reduction of valuable ecosystem services rangelands provide (Gudkaet al. , 2014; Kamp et al. , 2015). With the current land management policies and prediction of land-change we provide, the protected areas, especially Ghamashlu Wildlife Refuge, in our study area (figure 8) are at risk. For instance, our study revealed ongoing residential expansion in proximity to the Wildlife Refuge.
The presence of mineral sources in our study area, particularly in the mountains, resulted in the expansion of opened-cast mining, thus creating conflicts with environmental preservation strategies. Our land-change predictions and evaluation of existing plans about mining suggest further continuation of the negative impact of mining and supplementary road construction on rangelands and other drylands (Madadiet al. , 2014; Makki et al. , 2012). Expansion of big industrial towns and various mines over the last 30 years has also intervened between natural resource management and exploitation of water resources with other land uses such as farming (Soffianian & Madanian, 2015). Given a shortage of water resources in the study region of Iran, a vicious circle between the expansion of various land-use practices and water resources may result in further contraction of valuable land resources, and therefore, our results are extremely relevant for better land-use planning in the drylands of Iran.
In this study, we evaluated land-cover changes starting from 1985 to 2016. Land-change transformations started much earlier than our first year of observation (1985), for instance following the change of the political and institutional course in Iran in 1978. However, we captured an overall course of post-1978 land-cover change driven by the aim of satisfying the needs of the growing population. Before 1985, Landsat satellite imagery were scattered (Landsat 4 and 5 TM) or came with coarser spatial and radiometric resolution (Landsat 1-3 MSS). Similarly, we made projections about land-cover change by 2036 based on the overall course of land transformation in the past, therefore reflecting the ”business-as-usual” scenario. There might be other counterfactual scenarios that may mediate the rates and patterns of land-cover change, such as adjustment of existing land-use policies (Mallampalli et al. , 2016; Prishchepov et al. , 2019), change of economic conditions in Iran, etc. However, our projection can be used as a starting point of evaluation of existing policies and additional counterfactual scenarios can be built around.
Our study showed multiple pathways of land transformations in the dryland of central Iran, for instance, when both cropland expansion and abandonment concomitantly are taking place and go hand in hand with urban sprawl at the expense of rangelands. Despite the extreme climatic conditions, global drylands, which occupy almost half of the global surface (Prăvălie, 2016), are experiencing rapid land transformation due to the overall population increase. Moreover, a large portion of drylands are in developing countries, such as Iran, where livelihoods strongly depend on agriculture and environmental goods (Brandt et al. , 2017; Emg, 2011; Tong et al. , 2017). Despite a growing population across dryland, both cropland expansion and abandonment have also been found in other parts of the drylands of Northern Eurasia, such as Kazakhstan (Chen et al. , 2013; Horion et al. , 2016; Löwet al. , 2015), Mongolia (Dong et al. , 2011; Jamsranjavet al. , 2018; Sankey et al. , 2018) and Uzbekistan (Dubovyket al. , 2012). Considering the adverse climatic projections that drylands of Eurasia will face during the 21st century (Groisman et al. , 2017; Groisman & Soja, 2009), better land-use policies should be developed toward sustainable use of land and water resources to satisfy predicted population growth.
Acknowledgements
Authors thank DFF-Danish ERC Support Program (grant number: 116491, 9127-00001B) and the University of Copenhagen for a six month research stay at the University of Copenhagen, Department of Geosciences and Natural Resource Management (IGN).
References
Akbari N, Niksokhan MH, Ardesani M. 2015. Optimization of water allocation using cooperative game theory (Case study: Zayandehrud basin). Journal of Environmental Studies 40 : 875–890 and 10
Ansari A, Golabi MH. 2019. Prediction of spatial land use changes based on LCM in a GIS environment for Desert Wetlands – A case study: Meighan Wetland, Iran. International Soil and Water Conservation Research7 : 64–70. DOI: 10.1016/j.iswcr.2018.10.001
Azadi H, Barati AA, Rafiaani P, Raufirad V, Zarafshani K, Mamoorian M, Van Passel S, Lebailly P. 2016. Agricultural Land Conversion Drivers in Northeast Iran: Application of Structural Equation Model. Applied Spatial Analysis and Policy 9 : 591–609. DOI: 10.1007/s12061-015-9160-4
Behnke R, Robinson S, Milner-Gulland EJ. 2016. Governing open access: livestock distributions and institutional control in the Karakum Desert of Turkmenistan. Land Use Policy 52 : 103–119. DOI: 10.1016/j.landusepol.2015.12.012
Belgiu M, Drăguţ L. 2016. Random forest in remote sensing: A review of applications and future directions. ISPRS Journal of Photogrammetry and Remote Sensing 114 : 24–31. DOI: 10.1016/j.isprsjprs.2016.01.011
Brandt M, Rasmussen K, Peñuelas J, Tian F, Schurgers G, Verger A, Mertz O, Palmer JRB, Fensholt R. 2017. Human population growth offsets climate-driven increase in woody vegetation in sub-Saharan Africa.Nature Ecology & Evolution 1 . DOI: 10.1038/s41559-017-0081
Bürgi M, Hersperger A, Schneeberger N. 2004. Driving forces of landscape change — current and new directions. Landscape Ecology19 : 857–868
Castella J-C, Verburg PH. 2007. Combination of process-oriented and pattern-oriented models of land-use change in a mountain area of Vietnam. Ecological Modelling 202 : 410–420
Chaudhuri G, Clarke K. 2013. The SLEUTH land use change model: A review.International Journal Of Environmental Resource Research1 : 88–105
Chen T-HK, Prishchepov AV, Fensholt R, Sabel CE. 2019. Detecting and monitoring long-term landslides in urbanized areas with nighttime light data and multi-seasonal Landsat imagery across Taiwan from 1998 to 2017.Remote Sensing of Environment 225 : 317–327. DOI: 10.1016/j.rse.2019.03.013
Chen X, Bai J, Li X, Luo G, Li J, Li BL. 2013. Changes in land use/land cover and ecosystem services in Central Asia during 1990–2009.Current Opinion in Environmental Sustainability 5 : 116–127. DOI: 10.1016/j.cosust.2012.12.005
Cohen WB, Goward SN. 2004. Landsat’s Role in Ecological Applications of Remote Sensing. BioScience 54 : 535. DOI: 10.1641/0006-3568(2004)054[0535:LRIEAO]2.0.CO;2
Coppin P, Jonckheere I, Nackaerts K, Muys B, Lambin E. 2004. Digital change detection methods in ecosystem monitoring: A review.International Journal of Remote Sensing 25 : 1565–1596. DOI: 10.1080/0143116031000101675
Danaei G, Farzadfar F, Kelishadi R, Rashidian A, Rouhani OM, Ahmadnia S, Ahmadvand A, Arabi M, Ardalan A, Arhami M, Azizi MH, Bahadori M, Baumgartner J, Beheshtian A, Djalalinia S, Doshmangir L, Haghdoost AA, Haghshenas R, Hosseinpoor AR, Islami F, Kamangar F, Khalili D, Madani K, Masoumi-Asl H, Mazyaki A, Mirchi A, Moradi E, Nayernouri T, Niemeier D, Omidvari A-H, Peykari N, Pishgar F, Qorbani M, Rahimi K, Rahimi-Movaghar A, Tehrani FR, Rezaei N, Shahraz S, Takian A, Tootee A, Ezzati M, Jamshidi HR, Larijani B, Majdzadeh R, Malekzadeh R. 2019. Iran in transition. The Lancet 393 : 1984–2005. DOI: 10.1016/S0140-6736(18)33197-0
Dara A, Baumann M, Freitag M, Hölzel N, Hostert P, Kamp J, Müller D, Prishchepov AV, Kuemmerle T. 2020. Annual Landsat time series reveal post-Soviet changes in grazing pressure. Remote Sensing of Environment 239 : 111667. DOI: 10.1016/j.rse.2020.111667
De Rosa M, Knudsen MT, Hermansen JE. 2016. A comparison of Land Use Change models: challenges and future developments. Journal of Cleaner Production 113 : 183–193. DOI: 10.1016/j.jclepro.2015.11.097
Dong J, Liu J, Yan H, Tao F, Kuang W. 2011. Spatio-temporal pattern and rationality of land reclamation and cropland abandonment in mid-eastern Inner Mongolia of China in 1990–2005. Environmental Monitoring and Assessment 179 : 137–153. DOI: 10.1007/s10661-010-1724-9
Dubovyk O. 2017. The role of Remote Sensing in land degradation assessments: opportunities and challenges. European Journal of Remote Sensing 50 : 601–613. DOI: 10.1080/22797254.2017.1378926
Dubovyk O, Menz G, Conrad C, Kan E, Machwitz M, Khamzina A. 2012. Spatio-temporal analyses of cropland degradation in the irrigated lowlands of Uzbekistan using remote-sensing and logistic regression modeling. Environmental Monitoring and Assessment . DOI: 10.1007/s10661-012-2904-6
Eastman J. 2015. TerrSet: Geospatial monitoring and modeling software.