Environmental Challenges 4 (2021) 100199

Contents lists available at ScienceDirect

Environmental Challenges
journal homepage: www.elsevier.com/locate/envc

Agroforestry land suitability analysis in the Eastern Indian Himalayan

region
Arun Jyoti Nath a, Rakesh Kumar b,∗, N. Bijayalaxmi Devi c, Pebam Rocky d, Krishna Giri e,
Uttam Kumar Sahoo f, Raj Kumar Bajpai b, Netrananda Sahu g, Rajiv Pandey b
a
Assam University, Silchar 788011, Assam, India
b
Indian Council of Forestry Research and Education, Dehradun 248006, Uttarakhand, India
c
Sikkim University, Gangtok, Sikkim, India
d
North Eastern Space Application Centre, Umiam 793103, Meghalaya, India
e
Rain Forest Research Institute, Jorhat 785001, Assam, India
f
Mizoram University, Aizawal 796004, Mizoram, India
g
Delhi University, Delhi, India

a r t i c l e i n f o a b s t r a c t

Keywords: Indiscriminate deforestation and shifting cultivation have accelerated soil erosion requiring the need for immedi-
Multi-criteria decision approach (MCDA) ate conservation measures in the eastern Indian Himalayan region (EIHR). The present study attempts to address
Land management the problem by evaluating the land suitability of the EIHR for agroforestry through multi-criteria evaluation
Land suitability analysis
modelling through GIS. Several variables related to climate, soil, topography, ecology, and socioeconomic crite-
GIS
ria having high relevance for agroforestry were analysed and integrated to generate an agroforestry suitability
North-eastern region
map. The model output reveals that 60,523 sq. km (~77%) of the region’s arable land has very good to good
suitability for agroforestry, while 21,281 sq. km (~27%) area has very good suitability. The cross-validation by
utilising ground-based and literature-based information of existing agroforestry mostly supported the agroforestry
suitability map. The study concludes that agroforestry land suitability can be assessed through a multi-criteria
approach. The land suitability of the area can be enhanced further by suitable interventions such as improving
soil fertility, soil and water conservation.

1. Introduction and nutrient deficiency (Dagar and Tewari, 2017); with the limited re-
sources (Sileshi et al., 2014; Laskar et al., 2020).
The eastern Indian Himalayan region (EIHR) occupies 25 M ha and Agroforestry facilitates food security (MEA: Millennium Ecosys-
represents 8% of the total geographical area of India (IFSR: India State tem Assessment 2005) and poverty reduction (Mbow et al., 2014),
of Forest Report 2019). Indiscriminate deforestation and shifting culti- along with enhancing the resilience of households to climate change
vation (Jhum) are instrumental for the continuous degradation of eco- (Albrecht and Kandji, 2003). Additionally, agroforestry also im-
logical systems and soil in the region, producing adverse effects on the proves soil fertility (Ramos et al., 2015), conserve biodiversity
social systems (Nath et al., 2020; Laskar et al., 2020). The dominant (Hernandez et al., 2015), and improve the quality of the agroecosys-
soil in the EIHR is Acrisols and is strongly weathered acid soils with tems (Asbjornsen et al., 2014; NRCAF: National Research Centre for
low base (<50% base saturation) status and prone to erosion, particu- Agroforestry 2013). Therefore, India is intended to increase the total
larly on exposed sloping land (IUSS 2014). Around 30% of the area in area under agroforestry to 53 M ha by 2050 through restoring the fal-
the region has very severe soil erosion, with 60% of the area requires lows, cultivable fallows, pastures, groves, and rehabilitation of problem
erosion management (Mandal and Sharda, 2013). The steeply sloping soils (Dhyani et al., 2013) to make agroforestry a major land-use af-
landforms in the region support heavy runoff with an annual soil loss ter agriculture (140.86 M ha) and forestry (69.63 M ha) (Reisner et al.,
of more than 40 Mg ha−1 (Sharma, 2004). Agroforestry system can be 2007).
a potential intervention for rehabilitation and biological reclamation of Land Suitability Analysis (LSA) is a scientific method to assess the
problem soils such as degraded drylands soil prone to water and wind degree of fitness or suitability of land for a specific purpose (Singha and
erosion; acidic sulphate soils of humid regions characterized by low pH Swain, 2016; Ahmad et al., 2019; Hopkins, 1977) and aims to iden-
tify the most appropriate spatial pattern for future land uses according

∗
Corresponding author.
E-mail address: [email protected] (R. Kumar).

https://doi.org/10.1016/j.envc.2021.100199
Received 8 April 2021; Received in revised form 28 June 2021; Accepted 1 July 2021
2667-0100/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

to specific requirements, preferences, or predictors for intended pur- tropical, and warm tropical and high diversity of flora and fauna. The cli-
poses (Collins et al., 2001; Malczewski, 2004). LSA and Geographic In- mate of the EIH region is predominantly subtropical humid with hot and
formation System (GIS) are excellent tools for the sustainable planning humid summers, severe monsoons, and mild winters (Jhajharia et al.,
and management of the land (Collins et al., 2001; Malczewski, 2004; 2009) with annual mean temperature and humidity vary from 5 to 30 °C
Ahmad et al., 2020; Lopresti et al., 2015). It has been successfully and from 70% to 85%, respectively (Sharma et al., 2009). The EIH is
used to assess the suitability of land for specific crops (Herzberg et al., situated at the confluence of the Indo-Chinese, Indo-Malayan, and In-
2019; Elsheikh et al., 2013), agriculture (Zolekar and Bhagat, 2015; dian bio-geographical realms (Nath et al., 2019) and represents diverse
Aldababseh et al., 2018; Joss et al., 2008) and agroforestry (Singha and forests types as moist, dense, evergreen, semi-evergreen and temperate
Swain, 2016; Ahmad and Goparaju, 2017; FAO: Food and Agriculture forests (CEPF: Critical Ecosystem Partnership Fund 2005) with several
Organization of the United Nations 2007). The framework of land evalu- endemic and globally threatened species (Malczewski, 1999). The total
ation for agroforestry development requires the appropriateness of lands forest cover in the EIHR region is 1,70,541 sq. km., which is 65.05% of
in terms of the climate, soil, topography, environmental constraints, its geographical area (IFSR: India State of Forest Report 2019). The peo-
disturbances, socio-economic needs and anthropogenic pressure on the ple living in the proximity of forests are highly dependant on the forest
available lands (Pandey et al., 2017) besides the tree species and their for fuelwood, fodder, small timber, bamboo and many non-timber forest
management regimes (MEA: Millennium Ecosystem Assessment 2005). products.
These factors form a complex biophysical system comprising suitable
conditions and the constraints that determine the progress of any par- 2.2. Conceptual framework for assessing the suitable land for agroforestry
ticular agroforestry systems (AFS) on the land (Mbow et al., 2014). Such
an assessment for agroforestry would add value for appropriate planning The framework for LSA for agroforestry should be based on an inte-
to create nature-based solutions for mitigation and adaptation to climate grative and multi-disciplinary approach (FAO: Food and Agriculture Or-
change (Census of India 2011). ganization of the United Nations 2007) based on the criteria of climate,
LSA for agroforestry has been assessed at the country-level for soil, topography, ecology, and socioeconomic. The inputs derived from
India based on four parameters consisting of climatic (annual mean different datasets of each criterion have been reclassified (recoded),
temperature and precipitation), soil quality layer (nutrient availabil- ranked and integrated by the weighted overlay (WO) technique in GIS
ity and retention, oxygen availability to roots, root-penetrating capa- to produce the respective sub-models of agroforestry suitability of each
bility and toxicity), topographic (slope), and ecological criteria (Nor- criterion (Fig. 2).
malized Difference Vegetation Index (NDVI)) following FAO guidelines The WO approach was applied by assigning appropriate weights to
(Hopkins, 1977). The evaluation classifies 33% potential land of the each criterion based on their importance for agroforestry to integrate
country as highly suitable for agroforestry, with 40% moderately and them into a single output raster where more favourable factors result in
12% marginally relevant with the recommendation of region-based higher values and less favourable ones have lower values (Fig. 3). The
mapping to include region-specific characterization. However, due to WO is one of the most widely used approach in the decision-making pro-
the focus of the country-level assessment, the study lacks precision and cess in the domain of ecology and environment (Torabian et al., 2018;
not apt for local-level planning. Present evaluation attempts to pro- Tzioutzios and Kastridis, 2020), forestry (Sarkar et al., 2014), agricul-
vide information about the land suitability for agroforestry in the EIHR, ture (Basharat et al., 2016; Elsheikh et al., 2013; Joss et al., 2008) and
which is highly prone to land degradation. natural hazard management (Roslee et al., 2017; FAO/IIASA/ISRIC/ISS-
The specific objective of the study is the assessment of suitable land CAS/JRC 2012). The sub-models were integrated further by the WO
for agroforestry across the EIHR. It is hypothesized that agroforestry technique to generate the final model and maps of agroforestry suit-
suitability varies with the region’s climate, soil, topography, ecology, ability for EIHR (Fig. 3).
and socioeconomic factors. The regional level agroforestry suitability
mapping would provide an opportunity to policymakers for location- 2.3. Criteria and variables selected for agroforestry suitability assessment
specific land use planning for expanding and implementing agroforestry-
based models to address ecosystem restoration and food insecurity. A Five criteria, including climate, soil, topography, ecology, and so-
total of twenty-four sub-criteria, namely, annual mean temperature, an- cioeconomic, with their twenty-four variables (sub-criteria) were iden-
nual total precipitation, aridity index, and evapotranspiration for cli- tified for the present evaluation based on their influence and signifi-
mate; pH, organic carbon (OC), texture, cation exchange capacity (CEC), cance for agroforestry. The list of the datasets and variables related to
bulk density (BD) in topsoil and subsoil, available water capacity (AWC) the selected criteria and their nature and source are provided in Supple-
and drainage for soil; elevation, slope and aspect for the topography; mentary Material (Table 1).
NDVI and LULC for the ecology; and distance from the road, settlements,
and drainage network for the socio-economic criteria were considered to
2.3.1. Climate criteria and variables
classify and map the land suitability for agroforestry. A phased approach
The climatic condition of a region is an important determinant for
was adopted by separately integrating all variables of a criterion, and
the survival and growth of plants along with the nature and fertility of
finally, the sub-models of each criterion were integrated for obtaining
the soil (Hijmans et al., 2015). Plant growth is primarily dependant on
the final agroforestry suitability model.
light, temperature, and water (precipitation). Therefore, the following
four variables were considered under climate criteria for generating a
2. Materials and methods climate suitability sub-model of agroforestry suitability (Fig. 4).

2.1. Study area a) Temperature and Precipitation (P): The annual mean tempera-
ture and total annual precipitation of Worldclim version 2.1
The EIH region comprises Arunachal Pradesh, Assam, Meghalaya, (Farzanpour et al., 2019) for the period 1970–2000 were con-
Nagaland, Manipur, Mizoram, Tripura, Sikkim, Assam, and hill regions sidered.
of West Bengal and extends from 21° 56′ 26.296″ to 29°27′41.59″ North b) Reference Evapotranspiration (ETo): ETo is important for hydro-
latitude and 87° 59′ 14.024″ to 97° 24′ 43.345″ East longitude (Fig. 1). logical investigation and assessment of crop water requirements
The region spreads over 8.3% of the geographical area of India, with (Walter et al., 2000) and is estimated for an area using the
more than 456 million human population (Jain et al., 2012). The region Penman-Monteith equation based on the minimum and maxi-
is characterised by the highly rugged and complex mountainous topog- mum daily temperature, relative humidity, solar radiation, wind
raphy, with diverse climatic regions as arctic, sub-arctic, temperate, sub- speed, and latitude, and altitude of the area (Zomer et al., 2008).

2
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 1. Map of the study area: (a) shows the location of EIHR in the continents (b) shows the location of area in India and (c) shows the details of the EIHR region
with their physical and political setup.

c) Global Aridity Index (AI): AI, a measurement of the degree of dry- growth and development of plants. The slope of land determines the na-
ness of an area is typically a ratio of the long-term average wa- ture and the erodibility of the soil with high soil erosion in steep slopes.
ter supply (precipitation) to the long-term average atmospheric The GTOPO30 is a global digital elevation model (DEM) with a hori-
water demand (potential evapotranspiration) (Fischer et al., zontal grid spacing of 30 ′onds (about 1 km) and derived from several
2008). rasters and vector sources of topographic information (Didan, 2015).
The GTOPO30 DEM data was used for extracting information on eleva-
2.3.2. Topographic criteria and variables tion, slope and aspect of EIH (Table 1). The spatial layer of elevation,
The surface elevation, slope and aspect influence the ambient tem- slope and aspect (Fig. 5) were utilised for producing the topographic
perature, soil moisture and soil erosion which primarily affects the suitability sub-model for agroforestry.

3
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 2. Schematic representation of the methodology.

2.3.3. Soil criteria and variables of China, WISE) to the information contained within the FAO-UNESCO
Soil data for the EIH region was extracted from the Harmonised Soil Map of the World (Fischer et al., 2008). The global database has
World Soil Database (HWSD) developed by the Food and Agriculture over 16,000 different soil-mapping units at the raster grid size of 30
Organization of the United Nations (Hijmans et al., 2015) and the Inter- arc-second (about 1 km). Twelve soil variables (Fig. 6) relevant from
national Institute for Applied Systems Analysis (IIASA). This comprehen- HWSD for achieving different physical and biochemical requirements
sive global soil database has been prepared by combining updated infor- of the perennials plants and non-perennials crops of the agroforestry
mation on regional and national soil information (SOTER, ESD, Soil Map system was selected. They are texture, OC, pH, BD, CEC of the topsoil

4
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 3. Conceptual framework of weighted

overlay analysis.

Fig. 4. Spatial map for climatic variables: (a) Annual mean temperature; (b) Annual precipitation; (c) Reference Evapotranspiration and (d) Aridity Index.

(0–30 cm) as well as subsoil (30–100 cm), drainage and AWC of soil a) Nutrient availability in topsoil (SQtopsoil ): The pH, OC, texture, mi-
(Table 1). crobial population and total exchangeable base are important soil
Five soil quality (SQ) layers, described below, were prepared attributes for nutrient availability in soil. The spatial layer of top-
from the selected soil variables as per the Global Agroecological soil pH, OC, texture and CEC was integrated to generate a map
Zoning (GAEZ, v3.0) approach (Gomes et al., 2017). The inputs of the nutrient availability in topsoil.
(Fig. 6) of each soil quality layer were processed and integrated
through the WO method. The equal weight was assigned to each in- 𝑆 𝑄𝑡𝑜𝑝𝑠𝑜𝑖𝑙 = 𝑓 (𝑝𝐻, 𝑂𝐶, 𝑡𝑒𝑥𝑡𝑢𝑟𝑒, 𝐶𝐸𝐶 )
put for integration to generate the respective SQ layer. The five SQ
layers (Fig. 7) were utilised for producing soil suitability sub-model of b) Nutrient availability in subsoil (SQsubsoil ): The map of nutrient
agroforestry. availability in subsoil was generated by integrating pH, OC, tex-

5
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 5. Topographical variables: (a) Elevation (b) Slope and (c) Aspect.

ture and CEC of subsoil. 2.3.4. Ecological criteria and variables

The mean annual NDVI of 2019 and the forest cover data (IFSR: In-
𝑆 𝑄𝑠𝑢𝑏𝑠𝑜𝑖𝑙 = 𝑓 (𝑝𝐻, 𝑂𝐶, 𝑡𝑒𝑥𝑡𝑢𝑟𝑒, 𝐶𝐸𝐶 ) dia State of Forest Report 2019) have been considered under the eco-
logical criterion (Table 1, Fig. 8). The visible and near-infrared light
c) Rooting condition (SQrooting ): The pH and BD of topsoil and sub- reflected by the vegetation is used for calculating NDVI, which varies
soil were integrated to generate the spatial layers of the rooting from minus one (−1), indicating non-vegetated features, to plus one
condition of the soil. (+1), indicating vegetative features. The healthy vegetation, which ab-
( ) sorbs most of the incoming visible light and reflects a large portion of
𝑆 𝑄𝑟𝑜𝑜𝑡𝑖𝑛𝑔= 𝑓 𝑝𝐻𝑡𝑜𝑝𝑠𝑜𝑖𝑙 , 𝐵 𝐷𝑡𝑜𝑝𝑠𝑜𝑖𝑙 , 𝑝𝐻𝑠𝑢𝑏𝑠𝑜𝑖𝑙 , 𝐵 𝐷𝑠𝑢𝑏𝑠𝑜𝑖𝑙
the incoming near-infrared light, have a higher NDVI value. In con-
d) Oxygen Availability (SQoxygen ): The drainage characteristics of trast, the unhealthy or sparse vegetation, which reflects more visible
soils has a significant role in defining the oxygen availability in light and less near-infrared light, has a relatively lesser NDVI value. The
soils. Therefore, the soil drainage attributes from HWSD were NDVI is a dimensionless index where the values below 0.1 correspond
used for generating spatial layer of oxygen availability in soil. to bodies of water and bare ground. In contrast, higher values are in-
e) Water availability (SQwater ): The AWC attributes of HWSD was dicators of high photosynthetic activity linked to scrubland, temperate
used for generation of layer for water availability in soil. forest, rain forest and agricultural activity. The NDVI is used as an indi-

6
 A.J. Nath, R. Kumar, N.B. Devi et al.
Table 1
List of criteria, datasets and variables used for the agroforestry suitability analysis.

Criterion Name and source of the dataset Variables Unit Data Type Actual Grid size Resampled Grid size Source

Climate Standard World Clim historical ■ A nnual mean temperature °C Raster 30 s 1000 m (Farzanpour et al., 2019)
climate data (Average of ■ Total annual rainfall mm ( ̴ 1 km2 )
1970–2000)
Global Aridity Index ■ Global Aridity Index (Global-Aridity_ET0) unitless Raster 30 s 1000 m (Earth Resources
Evapotranspiration (ET0) Climate ■ Global Reference Evapotranspiration (Global-ET0) mm day−1 ( ̴ 1 km2 ) Observation and Science
Database v2. Center/U.S. Geological
(Average of 1970–2000) Survey/U.S. Department
of the Interior 1997)
Soil Harmonized World Soil Database ■ Bulk Density of topsoil and subsoil (kg/dm3 ) 30 s 1000 m (Hijmans et al., 2015)
(HWSD) V 1.2 ■ pH of topsoil and subsoil -log(H+ ) Raster ( ̴ 1 km2 )
■ Organic Carbon of topsoil and subsoil % weight
■ Texture of topsoil and subsoil code
■ Cation Exchange Capacity of topsoil and subsoil cmol kg−1
■ Available Water Capacity code
■ Soil Drainage code
code
7

Topography GTOPO30 Digital Elevation ■ elevation metre 30 s 1000 m (Didan, 2015)

Model ■ slope degree Raster ( ̴ 1 km2 )
■ aspect direction

Ecology Vegetation Indices Monthly L3 ■ Normalised Difference Vegetation Index Unitless 30 s 1000 m (ISCGM: Interna-
Global 1 km (November 2019) Raster ( ̴ 1 km2 ) tional Steering
Committee for Global
Mapping/Survey of India
2016)
Forest Cover map (2019) ■ Forest and Non-forest cover Code 24 m 1000 m (IFSR: India State of
Forest Report 2019)
Socioeconomic Global Map of India © ■ Built-Up Areas (Points), India, 2016 Vector (point) - - - (ISCGM: Interna-
ISCGM/Survey of India tional Steering
Committee for Global
Mapping/Survey of India
2016)

Environmental Challenges 4 (2021) 100199

defence mapping agency (dma), ■ Road network Vector (lines) - - - Defense Mapping
1992 ■ Drainage network Vector (lines) Agency (DMA) 1992
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 6. Topsoil: (a) bulk density, (b) organic carbon, (c) pH, (d) cation exchange capacity, (e) texture; Subsoil: (f) bulk density, (g) organic carbon, (h) pH, (i) cation
exchange capacity, (j) texture; (k) soil drainage, and (l) available water capacity of soil.

cator of the level of greenness in the area and widely used as the proxy availability and the restricted land for agroforestry purposes (Fig. 8).
for the quantity and quality of vegetation in terms of their health and The regions of non-forest, which include agricultural land, habitations
vigour (Effat and Hassan, 2013) and is an indicator of degradation of an and barren lands, scrub (canopy density less than 10%), and open for-
ecosystem. est classes (canopy density more than 10–40%), have been allocated for
The data on forest cover (canopy density more than 10%), non-forest the agroforestry suitability assessment. These regions have the potential
cover and waterbody were extracted from the India State of forest re- land where agroforestry interventions are needed to enhance productiv-
port, 2019 (IFSR: India State of Forest Report 2019) to determine the ity and ecosystem services or restore the degraded land. The areas un-

8
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 7. Five Soil quality inputs for soil suitability sub-models for agroforestry: (a) nutrient availability in topsoil (b) nutrient availability in subsoil (c) rooting
condition (d) oxygen availability and (e) water availability.

Fig. 8. Ecological variables: (a) NDVI (b) Forest and Non-forest cover.

9
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 9. Socioeconomic variables: (a) Road; (b) Distance from road; (c) Settlements; (d) Distance from settlements; (e) Drainage network and (f) Distance from
drainage network.

der moderate and very dense forest classes (canopy density more than 2.5. Weight assignment to the variables (sub-criteria) and criteria
40%) already have good vegetation cover with lesser land degradation
problems. Under this study, the area under permanent water bodies like Each input layer (variables) under each criterion has a unique
rivers and lakes has been kept under a restricted class for agroforestry property that determines its importance for agroforestry. Depending
interventions. upon the relative significance of each variable/criteria, their respec-
tive weights are determined by several subjective and objective meth-
2.3.5. Socioeconomic criteria and variables ods (Saaty and Vargas, 2013). In this case, a direct weight assign-
Socioeconomic criteria and their variables were included for a com- ment method based on the domain expert’s understanding of the signifi-
prehensive LSA for agroforestry. The vector data on the existing road cance of criteria was followed (Singha and Swain, 2016; Hopkins, 1977;
network, locations of settlements, and the drainage network have been Saaty and Vargas, 2013). While equal weights assigned to each variable
used in this study (Fig. 9). Euclidean distance operator is applied to the under the soil, socioeconomic and ecological criteria, unequal weights
data for producing the raster distance file for each of the considered were provided to variables of climate and topographic criteria. Like-
variable. wise, the final weights of each criterion were also based on the domain
expert’s knowledge.
2.4. Standardization and ranking of each variable of criteria The specific weights assigned to each variable and each criterion is
provided in Table 2. The variable of higher significance was provided
The variables under different criteria have different nature and di- with higher weights. The respective weights of each raster were multi-
mensions and are reclassified to achieve uniformity for their compari- plied by each cell value of that raster. The weight to each variable under
son and integration in GIS by the reclassification method (Dragan et al., a criterion is provided in such a way so that the sum of all the weights
2018). All the variables were transformed to a scale of 1 to 5, with "1″ was 100%. The standardized variables of each criterion were integrated
as the highest, and "5″ as the lowest suitability for agroforestry. The dis- using the WO procedure in GIS to obtain the agroforestry suitability
tance variables (distance from road, settlements, and drainage network) sub-model for each of the five criteria.
were also standardized and ranked by a similar method with rank ’1′ to
the nearest space as highest and rank ’5′ to the farthest as lowest suit- 2.6. Generation of agroforestry suitability model
ability.The higher suitability rank (1st or 2nd) was assigned to the most
suitable range of each criterion’s variables. In contrast, lower (3rd or The final suitability model of agroforestry was generated by inte-
4th) to lowest (5th) rank was assigned to a less suitable or currently un- grating the agroforestry suitability sub-model of climate, topography,
suitable range of each criterion’s variables (Table 2). The suitable scores soil, ecology and socioeconomic through the WO method. The weights
(ranks/classes) to different ranges of each variable under each criterion assigned (Table 2) to each criterion suitability sub-model were based on
were decided based on the relevant available literature, discussion with their relative significance and contribution to agroforestry suitability.
peers, and knowledge of the domain experts. A sensitivity analysis of the final agroforestry suitability model was

10
 A.J. Nath, R. Kumar, N.B. Devi et al.
Table 2
Suitability classes and weights assigned to variables and respective criteria as per the available literature and knowledge of domain experts.

Criteria (c) Variable (v) Unit of Range/Code Suitability classes for agroforestry (1: Very good, 2: Good, 3: Moderate, 4: Marginal, 5: Weight for Weight for
measurement Unsuitable) variable (wv ) criteria (wc)

1 2 3 4 5

Climate Annual mean temperature °C -19.5 to 25.5 20-25.5 15-20 10-15 5-10 <5 35% 35%
Total annual precipitation mm 235 to 9312 >2000 2000-1500 1500-1300 1300-1000 <1000 35%
Reference Evapotranspiration mm/day 73-1663 1450-1663 1250-1450 1100-1250 1000-1100 <1000 15%
Aridity Index unitless 2442-70313 35000> 15000-35000 10000-15000 5000-10000 <5000 15%
Soil Topsoil pH Nutrient -log(H+ ) 4.6 to 7.6 5.5-7.2 7.2-7.4; 7.4-7.6 - - 25% 20% 20%
availability in 4.6-5.5
Organic carbon topsoil % weight 0.56 to 3.71 2.0-3.71 1.0-2.0 0.6-1.0 <0.6 25%
Texture code Coarse to fine fine medium coarse - - 25%
Cation exchange cm/kg 4-26 3-23 >10 5-10 - - 25%
capacity
Subsoil pH Nutrient -log(H+ ) 4.7 to 7.5 5.5-7.2 7.2-7.4; 7.4-7.6 - - 25% 20%
availability in 4.7-5.5
Organic carbon subsoil % weight 2 to 1.03 0.6-1.03 0.4-0.6 0.2-0.4 <0.2 - 25%
Texture code 1-12 5, 7, 9 4,12 1,3,10, - - 25%
11

Cation exchange cm/kg 3-23 >10 5-10 <5 - - 25%

capacity
Topsoil pH Rooting -log(H+ ) 4.6 to 7.6 5.5-7.2 7.2-7.4; 7.4-7.6 - - 25% 20%
condition 4.6-5.5
Bulk density Kg/dm 3
1.1-1.52 <1.40 1.40-1.44 1.44-1.48 >1.48 - 25%
Subsoil pH -log(H+ ) 4.7 to 7.5 5.5-7.2 7.2-7.4; 7.4-7.6 - - 25%
4.7-5.5
Bulk density Kg/dm3 1.16-1.66 <1.2 1.2-1.4 1.4-1.58 1.58-1.66 - 25%
Soil Drainage Code 1,2,3,4,6 4 3,6 2 1 - - 20%
Available water capacity Code 1, 3, 5 6, and 7 1 3 5 6 7 - 20%
Topography Elevation m 3 to 7496 3-1500 1500-2200 2200-3500 3500-4500 4500-7496 40% 25%
Slope degree 0 to 44 0 to 5 5 to 10 10 to 15 15 to 20 20 to 44 40%
Aspect Direction N, S, E and W S, E , W - N - - 20%
Ecology NDVI Unitless 0 to 1 0.6-0.9 0.4-0.6 0.2-0.4 0.1-0.2 >0.1 50% 10%
Forest Cover map Category Water, NF, Non-forest, - - - Water, MDF & 50%
Scrub, Forest Scrub, OF VDF
Socioeconomic Distance from Roads m 0-43140 0-3 3-6 6-10 10-20 >20 34% 10%

Environmental Challenges 4 (2021) 100199

Distance from Settlements m 0-77897 0-5 5-10 10-15 15-25 >25 33%
Distance from Drainage m 0-51855 0-2 2-4 4-7 7-10 >10 33%

Note: The weights for variable and criterion are as per literature and expert. NF: Non-forest; OF: Open forest.
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 10. Climate suitability map of agroforestry.

performed to determine the relative strength and practicality of each 2.7. Ground validation of agroforestry suitability model
criterion for final suitability. It aids in defining the level of significance
of each criterion while also help in removing the subjectivity factor The final agroforestry suitability map was validated through field-
in the assignment of weights to them (Jain, 2014). To undertake this based location information of actual agroforestry practice in the EIHR.
analysis, the initial input weight of selected criteria was changed while Due to the enormous size and terrain complexity of the study area,
keeping the weights of other criteria unchanged to see the effect of that it was not feasible to validate the whole area through a probability
change on the model output. As any alteration observed in the output (random) sampling method. Therefore, we have selected two zones
was explicitly happened due to the new weights assigned to the single one from the northern (Brahmaputra valley) and the other from the
input criteria. It is a rational approach to assess the overall priority southern (Barak valley region) and selected some random locations
of each input criteria with respect to others and help in objectively of existing agroforestry and noted the geo-coordinates from the field.
comparing the model outputs (Joss et al., 2008). The sensitivity of each The ground-based agroforestry locations were processed in GIS and
criterion was analysed through this method to ascertain their relative overlaid on the agroforestry suitability map. The accuracy of the final
significance for agroforestry. suitability map was determined by comparing the model output with
The agroforestry suitability analysis was carried for all the land with the corresponding actual agroforestry information collected from the
non-forest, scrub and open forest cover only. However, the land under field.
permanent water and moderately to very dense forest cover was kept Moreover, a literature-based approach was also utilised to verify the
under restricted class and not included in the final suitability evalua- agroforestry suitability classes by comparing the relevant information
tion. The suitable area for agroforestry was also analysed for its spatial such as agroforestry type along with their location coordinates derived
distribution on the arable land existing in different states of the EIH from the published literature on agroforestry relevant to the study
region. region.

12
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 11. Topographic suitability map of agroforestry.

3. Results As per the climate suitability map for agroforestry (Fig. 10), about
half of the region has highly suitable (very good) climate condition for
3.1. Climatic criteria and its suitability for agroforestry agroforestry with about 32% and 6% of the region has a good and mod-
erately suitable climate for agroforestry, respectively. The rest of the
The status and variability of the annual mean temperature, total pre- area of the region located at a very high altitude has either a marginally
cipitation, evapotranspiration and aridity condition prevailing in the suitable or an unsuitable climatic condition for agroforestry.
EIHR are presented in Fig. 4. The annual temperature varies from −19.5
to 25.5 in the EIHR. The high mountain regions with higher altitude 3.2. Topographical criteria and its suitability for agroforestry
(>1500 m) have a lower mean temperature ranging from 2°−15°, and
lower altitude regions (<1500 m) has a higher (15°- 25.5°) range of an- The spatial distribution and pattern of elevation, slope and aspect
nual mean temperature and receive the highest amount of rainfall in prevailing over the EIHR is provided in Fig. 5. The whole region has
eastern Himalaya. The majority (63%) of the area at the lower eleva- mostly mountainous topography with elevation ranging from low (3–
tion region receives high (2000–3500 mm) to very high (>3500 mm) 500 m) to moderate (500–1500 m) and high (1500–3500 m) to very high
annual precipitation, and higher altitude regions mainly in the North (>3500 m). More than two-thirds (76.6%) of the geographical region has
receive low annual precipitation. Moreover, most locations in mid to an elevation below 1500 m, and around 18% has a high (1500–3500)
lower altitudes had moderate to high evapotranspiration. However, the elevation. Despite the predominantly mountainous region, around 57%
high mountain region and snow cover had low evapotranspiration. The of the region has a flat topography with a very gentle slope (<2°) with
Meghalaya state has the highest aridity index in the region, i.e. the around 12% gentle (2 to 5°), and 22% undulating slopes (5 to 8°). The
moistest in the EIHR. The areas with high to very high evapotranspi- 5% of the region has a rolling slope (8 to 16°), and 4% hilly to steep (16
ration are comparatively less humid and moderately dry. to 45°) slopes. The region has a very equitable distribution of terrain

13
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 12. Map of soil suitability of agroforestry.

aspect. 28.6% of the area has a southern aspect, while around 21.4% EIHR has good nutrient in the topsoil and central EIHR has moderate
has a northern aspect. The eastern and western aspect has almost equal nutrient in the topsoil. The subsoil in the southern EIHR region com-
distribution with 24.6% and 25.4%, respectively. prising Nagaland, Manipur, Mizoram, plains of Tripura and Arunachal
The regions with moderate elevation in conjunction with gentle slope Pradesh, western Meghalaya and southern and central plain of Assam
are the favourable topography for agroforestry. About 58,695 sq. km have high to a good level of the nutrient. The majority of the north-
(21%) of EIHR, mostly distributed over the Brahmaputra and Barak ern part of EIHR (Hilly region of Arunachal Pradesh and West Bengal
valley of Assam, western Meghalaya, plains of Tripura, Mizoram, Na- and Brahmaputra valley of Assam) has moderate to low nutrients in the
galand and Manipur have highly suitable (very good) topography for subsoil.
agroforestry. Around 1,32,052 sq. km (48%) area of the region’s to- The southeastern region of the EIHR region has highly suitable soil
pographic condition is good for agroforestry, while about 29% has for plants rooting supports, however, the majority of the remaining area
moderate to marginally suitable topography for agroforestry purposes of EIHR has good to moderately suitable soil for plants rooting supports.
(Fig. 11). The majority of the area of the EIHR region has a high to a good level
of oxygen availability in soil due to optimum drainage in the soil. Only
3.3. Soil criteria and its suitability for agroforestry some parts of southern Barak valley, eastern Meghalaya, southern hilly
region of West Bengal has moderate to low oxygen availability due to
The condition of the soil was evaluated in terms of availability of nu- poor drainage condition in the soil. Water availability in soil was found
trient, oxygen, water and rooting condition for the optimum growth of optimum to support plants in most of the northern, central and western
plants (Fig. 6). The soil of the western and southern Meghalaya, Barak parts of the EIHR. The western parts of Mizoram and the southwest-
valley, eastern and western Manipur, western part Mizoram and Na- ern part of the EIHR has a moderate level of water availability in soil
galand region of EIHR has high nutrients in topsoil, while northern (Fig. 6).

14
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 13. Map of ecological suitability of agroforestry.

All the five soil quality variables were integrated to form the soil The NDVI and land cover variables were integrated to generate the
suitability map for agroforestry (Fig. 7). The analysis of this map reveals ecological suitability map for agroforestry (Fig. 13). Only the region
that the majority (84%) of EIHR has suitable soil with very good to good with non-forest, scrub and open forest type land cover has been anal-
potential for agroforestry intervention. However, the remaining area has ysed for the suitability for the agroforestry. Out of vegetated area, about
mostly relatively poor quality soil with moderate to very low suitability 23% of the land was highly favourable and 71% had a good ecological
for agroforestry (Fig. 12). condition for agroforestry.

3.4. Ecological criteria and its suitability for agroforestry 3.5. Socioeconomic criteria and its suitability for agroforestry

The ecological status of the EIHR was evaluated based on NDVI and The spatial distribution of distance from the human settlements, road
land cover type provided in Fig. 8. The region has a tropical to sub- network and drainage network was analysed for agroforestry suitability
tropical climate with very humid to semi-humid conditions (Nair, 2011) (Fig. 9). The human settlements were well distributed throughout the
highly suitable for supporting luxuriant forest and vegetation. The area region, with a higher concentration in low altitude with gentle slopes.
with very dense forest cover has high (0.6–0.9) NDVI, and the region Only 10% of the area was found within 3 km distance, about 53% is
with moderately dense vegetation has moderate (0.4–0.6) NDVI. The found located within 10 km from the nearby settlement’s location. The
area with mostly agriculture-dominated landscapes or sparse and open road infrastructure of the region is very poor. Most of the rural area has
vegetation has lower (0.2–0.4) NDVI. The region is endowed with rich poor accessibility. About 36% and 33% of the area was found within
forest and vegetation on 65% of geographical area with 39% of the to- 3 km and 7 km of some nearby road infrastructure, respectively. The
tal area with moderate to the very dense forest, while 26% area under majority of the area was located close to nearby drainage. About 31%
open forest. Agriculture, grassland, herbaceous vegetation, waterbody, area of the region found within 2 km and 35% area within 2 to 4 km
settlements, and snow cover dominate the remaining area. distance of the drainage network (Fig. 9).

15
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 14. Map of socioeconomic suitability of agroforestry.

The socioeconomic suitability map of agroforestry was generated by 22,641 sq. km (36%), 6966 sq. km (47%) and 4565 sq. km (30%), re-
integrating the three standardized distance variables of socioeconomic spectively. However, the majority of states of the EIHR have more than
criteria. Only 26% of the EIH region has a highly favourable location 60% of available land with relatively good suitability for agroforestry
(within 5 km aerial distance), while 54% is within 10 km of either any (Table 3). While the relatively less hilly region of Sikkim and Arunachal
road or habitation of any source stream (drainage). About 16% of the Pradesh had high suitability, in contrast, the higher hilly regions of these
area is 10–15 km away from such a landmark (Fig. 14). states had only moderate to the marginal level of suitability for agro-
forestry (Table 1).
3.6. Agroforestry suitability of EIHR

The five sub-models of agroforestry suitability for each criterion were 3.7. Agroforestry suitability of arable land of EIHR
integrated to generate the final suitability for agroforestry in EIHR. The
land area with non-forest, scrubs and open forest cover has only been The final output of agroforestry suitability was also analysed for its
considered for agroforestry suitability assessment. About 29% (47992 spatial distribution in the arable land available in EIHR (Fig. 16). The
sq. km) of the land area in EIHR was found highly suitable (very good) analysis reveals that more than two-third (~77%) of all arable land of
and about 60% (98,236 sq. km) is found good for agroforestry purposes. the region are either very good or good in terms of suitability for agro-
The area with a moderate level of agroforestry suitability was about 6%. forestry. While 21,281 sq. km (~27%) area has very good suitability,
The remaining area with perennial snow is currently not suitable for about 39,242 sq. km (~ 50%) area has been found good for agroforestry.
agroforestry (Fig. 15). The region of Assam, Manipur, Mizoram and hill districts of West
Assam, Mizoram and Manipur were the leading states in terms of to- Bengal have the maximum area under the arable land with either a very
tal area with very good suitability for agroforestry with a land area of good or good level of suitability for agroforestry. The state-wise statistics

16
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 15. Agroforestry suitability map of Eastern Indian Himalayan region.

Table 3
State-wise agroforestry suitability area for EIH region (sq. km).

State Allocated land∗ Very good % Good % Moderate % Unsuitable %

Arunachal Pradesh 30,140 3636 12 11,969 40 8451 28 6028 20

Assam 62,903 22,641 36 40,043 64 – – – –
Manipur 14,971 4565 30 10,489 70 12 – – –
Meghalaya 12,472 3552 28 8980 72 – – – –
Mizoram 14,828 6966 47 7859 53 – – – –
Nagaland 10,657 1833 17 8817 83 28 - – –
Sikkim 4295 194 5 570 13 1321 31 2169 51
Tripura 4254 1524 36 2723 64 – – – –
West Bengal 9871 3081 31 6786 69 2 – – –
Total 1,64,391 47,992 29 98236 60 9814 6 7859 5
∗
Allocated land: Includes non-forest, scrub and open forest land in each state of the EIH region considered for final agroforestry suitability evaluation.

of agroforestry suitability of arable land of each state is provided in The GPS locations of agroforestry points were matched with correspond-
Table 4. ing locations in the agroforestry suitability map. Out of all the ground-
based locations, 56 points (about 86%) have supported the model out-
3.8. Validation and sensitivity analysis of agroforestry suitability mapping put and the remaining locations were found to be located under the
restricted class. These nine points were found to be located in the re-
GPS-based ground coordinates of 65 random locations of actual agro- gion of moderately to dense forest cover area that has been kept as the
forestry collected from the two selected zones of the EIH region (Fig. 16). restricted area and not included in the final agroforestry suitability map

17
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Fig. 16. Map of agroforestry suitability of arable land of EIHR along with the GPS locations for verification.

Table 4
State-wise agroforestry suitability area (sq. km) of arable land.

State Arable land∗ Very good % Good % Moderate % Unsuitable %

Arunachal Pradesh 11,596 1091 9 2456 21 2348 20 1312 11

Assam 40,764 13,703 34 23,333 57 – – – –
Manipur 3074 924 30 1704 55 2 – – –
Meghalaya 4290 968 23 1797 42 – – – –
Mizoram 2624 808 31 1151 44 – – – –
Nagaland 4690 811 17 2771 59 6 – – –
Sikkim 1385 115 8 255 18 301 22 235 17
Tripura 3034 562 19 1105 36 – – – –
West Bengal 7514 2299 31 4670 62 – – – –
Total 78,971 21,281 27 39,242 50 2657 3 1547 2
∗
includes cropland, fallow land, shrubland, grassland, plantations and wasteland.

of the EIH region. The result of the class-wise comparison of ground information of agroforestry types and their ground coordinates were ex-
control points with model-derived agroforestry suitability classes is pre- tracted wherever available from the literature. All the literature-based
sented in Table 5. We have analysed the sensitivity of each criterion by reported locations of the existing agroforestry across the EIHR also
altering the weights of individual criteria, one at a time, and found that found to be matching with model-derived agroforestry suitability classes
the climate and topography play the most significant role followed by (Supplementary Material, Table S1).
topography, soil, ecology and socioeconomic criteria in influencing and
determining land suitability for agroforestry. 4. Discussion
The model derived agroforestry locations were also compared with
agroforestry locations derived from the published peer-reviewed liter- Agroforestry is an ancient practice that combines trees with shrubs,
ature related to agroforestry practices of the EIH regions. The relevant crops, and livestock in a system that produces food, supports biodiver-

18
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Table 5
Comparison of ground-based agroforestry locations with model-derived class.

Agroforestry system (ground) Model-derived agroforestry suitability class

Very good Good Restricted Total Matching%

Home gardens 37 13 5 55 91
Piper-beetle agroforestry 0 1 3 4 25
Fruit tree-based agroforestry 1 1 0 2 100
Bamboo-based agroforestry 1 0 0 1 100
Eri-silk agroforestry system 1 0 0 1 100
Ananas-based agroforestry 0 1 1 2 50

sity, builds soil horizons and water tables, and sequesters carbon from criteria decision approach (MCDA) with GIS application enables this
the atmosphere (Gebre, 2016; Agroforestry module 2021). Although evaluation convenient and quick to manage (Yen et al., 2013).
practised worldwide, its nature and distribution varied from country to The MCDA assists in achieving more accurate results using more cri-
regions and governed by human needs and capabilities and the prevail- teria and finer scales data (Fischer et al., 2014). However, integrat-
ing environmental, cultural and socioeconomic conditions. Some of the ing several criteria (variables) into a single suitability model makes
popular agroforestry systems worldwide are Taungya; home gardens; the model complicated and relatively difficult to manage. The rela-
alley cropping; silvopastoral systems; growing multipurpose trees and tive significance of each variable and each criterion of plants sur-
shrubs in farmlands; boundary planting; farm woodlots; orchards and vival and growth are the core factor for assigning weights (Singha and
tree gardens; tree plantations; shelterbelts; windbreaks; conservation Swain, 2016; Hopkins, 1977). However, detailed and site-specific data
hedges; fodder banks; live fences; and apiculture with trees (IPCC 2019). of each variable of relevant criteria are recommended for farm level or
It is estimated that the area currently under agroforestry worldwide species-specific suitability modelling (Hancock et al., 2011). Most of the
is 1023 M ha. As per an estimate by Intergovernmental Panel on Cli- variables chosen under each criterion have been on the same scale (1 km
mate Change (IPCC), about 630 M ha areas of unproductive croplands, spatial resolution). Therefore, the approach was without constraint from
grasslands, and degraded lands could be brought under agroforestry the upscaling or downscaling of the data. All the variables used for the
(Chavan et al., 2015). As agroforestry is a more complex system than landscape-level analysis had good quality precision, except the soil vari-
a monoculture, it requires considerable effort, time, and technical ex- ables from HWSD, as HWSD data of the South Asian regions are of com-
pertise to successfully adopt a suitable agroforestry system in potential paratively low quality (Hijmans et al., 2015).
regions. Simulation models and decision-making tools such as MCDA Agroforestry systems across the EIHR have a wide variety depending
based land suitability analysis similar to this study have proven their on climatic, soil and topographical aspects (Supplementary data: Table
high value in determining the most suitable agroforestry system based S1). Agroforestry systems exist across the EIHR, ranging from the humid
on the needs and aspirations of land users along with prevailing eco- tropical lowlands of Assam and Tripura to the high-altitude and temper-
nomic, social and environmental conditions of the region and land- ate climate of Sikkim and covering diverse eco-zones of the region. Sim-
scape. Moreover, it can also help refine the efficiency of the existing ilarly, sub-tropical regions of Arunachal Pradesh, Manipur, Mizoram,
agroforestry system by evaluating the existing practices and exploring Nagaland and Meghalaya also support different forms of agroforestry
possible improvements to them. (Table S1). The locations of existing agroforestry systems reported in
India has a national agroforestry policy that supports smallholder the published literature fall well within the agroforestry suitability map
farmers (Dhyani, 2014; Purevtseren et al., 2017) and provides opportu- developed in the study, suggesting the agroforestry suitability modelling
nities for scaling up agroforestry practice in the EIHR. However, the approach (Table S1).
selection of suitable agroforestry systems (for example, agrisilvicul- Climatic factors such as optimum temperature and moisture condi-
ture/silvopasture etc.), for the different region may vary with climatic tion are vital for tree species distribution and adaptation (FAO 2008)
and local environment condition along with the type the tree species which highly influence and determine their climate suitability and hence
to be used. The approach utilised in this study can be instrumental in productivity (Rana et al., 2018). Our study indicates that the humid sub-
making decisions related to selecting a suitable agroforestry system at a tropical climate of the EIHR, which covers around 50% of the region,
regional or local scale. This study has analysed the climate, topography, is highly suitable for agroforestry. The study also revealed that about
soil, land cover, and socioeconomic criteria for evaluating the potential 1,32,052 km2, i.e. 48% of this region’s topographic condition, is good
region for agroforestry intervention and expansion. Overall, about 29% for agroforestry, while about 29% has moderate to marginally suitable
(47,992 sq km) of the land under non-forest, scrubs, and open forest topography for agroforestry purposes. Agroforestry evaluation carried
cover in the EIHR were found highly suitable. About 60% (98,236 sq in other regions also realised the high significance of climatic factor
km) have good suitability for agroforestry purposes. Assam, Mizoram and local environmental conditions that contributes to the successful
and Manipur were the central states in terms of total area with high development of agroforestry systems (Szott et al., 1991).
suitability for agroforestry with land area 22,641 sq. km (36%), 6966 The potential for agroforestry systems is often constrained by the
sq. km (47%) and 4565 sq. km (30%), respectively (Table 3). lack of suitable soil properties (Brahma et al., 2018). Large-scale defor-
Land characteristics, such as topography, soil fertility, drainage, wa- estation and shifting cultivation are the prime cause of soil degradation
ter resources, climate and environmental conditions, such as precipita- in the EIHR (Laskar et al., 2020). In the EIHR, following large-scale
tion, temperature, and seasonal variations, need to be considered in the logging, much of the previous rainforest areas have been converted to
decision-making process for land suitability evaluation for agroforestry large-scale monocultures rubber (Hevea brasiliensis) for synthetic rub-
development (Pandey et al., 2017). A multi-disciplinary approach inte- ber and Areca plantations for betel nut production (Das et al., 2021;
grating the biophysical, economic, social, political, and environmental Nath et al., 2018). Conversion of forests to monoculture plantations has
concerns has been recommended for land evaluation (Elsheikh et al., been shown to result in a significant decline in ecosystem carbon stocks
2013; Zolekar and Bhagat, 2015). In the present study, the land suit- (Das et al., 2021). Such land-use conversion to monoculture plantations
ability evaluation was performed by integrating many contextual crite- has often been associated with various environmental issues, includ-
ria such as climate, soil, topography, ecology, and the socioeconomic ing biodiversity loss (Nath et al., 2021). To address the problem of soil
attributes of the mountainous landscape of EIH. The use of the multi- degradation and decline in productivity, a suitable agroforestry system

19
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

can be intervened in the affected region following the approach of this Declaration of Competing Interest
study. Our study found that 84% of the EIHR has suitable soil with high
to good potential for agroforestry intervention. Our study also suggested The authors declare that they have no conflict of interest.
that about 23% of the land has high and 71% has good ecological suit-
ability for the agroforestry system in the eastern Himalayan region. Author contributions
Agroforestry systems, an essential land-use in India covering 25.3
million (Dhyani, 2014; Purevtseren et al., 2017; IFSR: India State of RP, UKS, AJN, and RK conceived the ideas. RP, AJN and RK designed
Forest Report 2019), has a potential for mitigation and adaptation to methodology. RP, BD, KG, NS and RK collected the data; RP and RK
climate change (Census of India 2011; Press Information Bureau 2019). analysed the data; RP, AJN, NS and RK led the writing of the manuscript.
India is committed to restoring 26 million ha of degraded land by All authors contributed critically to the drafts and gave final approval
2030 as per the United Nations Convention to Combat Desertification for publication.
(Sahoo, 2009). In the EIHR, agroforestry is well-positioned to tackle land
degradation on cropland while also creating a C sink. The expansion of
Acknowledgements
a suitable agroforestry model in the cultivable and degraded land in the
EIHR can offer several protective and productive services. Promotion
Financial assistance received from the Department of Science and
and support to a need-based locally suitable agroforestry system will en-
Technology under the research grant DST/CCP/MRDP/189/2019 is
hance livelihood, support local biodiversity, improve the environment,
highly acknowledged.
and mitigate climate change (FAO 2017; Mishra et al., 2018).
Commercial agroforestry is gaining momentum due to large-scale
demand from India’s pulp and paper industries (Dhyani, 2014). More- Supplementary materials
over, the people of the EIHR practised location-specific diverse tradi-
tional agroforestry systems since time immemorial. For example, the Supplementary material associated with this article can be found, in
alder-based systems, a common agroforestry practice in some parts of the online version, at doi:10.1016/j.envc.2021.100199.
the EIHR been reported to improve soil nutrient contents in Nagaland’s
depleted soil [82]. Traditional home gardens and Piper-betel based agro- References
forestry in the EIHR have sustained the people’s livelihood for millennia
Agroforestry module, 2021. Sustainable Forest Management (SFM) Toolbox.
(Press Information Bureau 2019). Tea is often grown under a canopy FAO http://www.fao.org/sustainable-forest-management/toolbox/modules/
of trees, forming a unique agroforestry system covering a total area of agroforestry/basic-knowledge/en/.
about 0.5 M ha in the EIHR (Press Information Bureau 2019). Ahmad, F., Goparaju, L., 2017. Geospatial approach for agroforestry suitability mapping:
to enhance livelihood and reduce poverty, FAO based documented procedure (Case
The coarse datasets with a spatial resolution of 1 sq. km have been Study of Dumka District, Jharkhand, India). Biosciences Biotechnol. Res. Asia 14 (2),
mostly considered in this study for generating the final agroforestry suit- 651–665. doi:10.13005/bbra/2491.
ability map without targeting any specific species or agroforestry model. Ahmad, F., Uddin, M.M., Goparaju, L., 2019. Agroforestry suitability mapping of India:
a geospatial approach based on FAO guidelines. Agrofor Syst 93 (4), 1319–1336.
Therefore, the outputs may not be suitable for the finer-scale local site doi:10.1007/s1045-7-018-0233-7.
due to their specific microclimatic condition. Therefore, Site-specific de- Ahmad, F., Uddin, M.M., Goparaju, L., et al., 2020. Quantification of the Land Poten-
tailed information on biophysical and environmental condition based tial for Scaling Agroforestry in South Asia. KN J. Cartogr. Geogr. Inf. 70, 71–89.
doi:10.1007/s42489-020-00045-0.
on desired agroforestry is recommended for finer scale suitability anal- Albrecht, A., Kandji, S.T., 2003. Carbon sequestration in tropical agroforestry systems.
ysis with higher precision output. The inclusion of stakeholders’ pref- Agric. Ecosyst. Environ 99, 15–27.
erences and inputs in this process will make such exercise more fruit- Aldababseh, A., Temimi, M., Maghelal, P., Branch, O., Wulfmeyer, V., 2018. Multi-criteria
evaluation of irrigated agriculture suitability to achieve food security in an arid envi-
ful and practicable. The site-specific suitable agroforestry design can be
ronment. Sustainability 10 (3), 803. doi:10.3390/su10030803.
achieved through our land evaluation approach to reclaim abandoned Asbjornsen, H., Hernandez-Santana, V., Liebman, M., Bayala, J., Chen, J., Helmers, M.,
or degraded shifting agriculture lands. The suitable mixture of crops and Schulte, L., 2014. Targeting perennial vegetation in agricultural landscapes for
enhancing ecosystem services. Renewable Agric. Food Syst. 29 (2), 101–125.
tree species to enhance productivity and the sustainable developments
doi:10.1017/S1742170512000385.
of the local community can also be realised by the scientific land eval- Basharat, M., Shah, H.R., Hameed, N., 2016. Landslide susceptibility mapping using GIS
uation process similar to this study. and weighted overlay method: a case study from NW Himalayas. Pakistan. Arab J
Geosci 9, 292. doi:10.1007/s12517-016-2308-y).
Brahma, B., Nath, A.J., Sileshi, G.W., Das, A.K., 2018. Estimating biomass stocks and
potential loss of biomass carbon through clear-felling of rubber plantations. Biomass
5. Conclusions Bioenergy 115, 88–96. doi:10.1016/j.biombioe.2018.04.019.
Census of India, 2011. Primary Census Abstract Highlights. Office of the
Registrar General & Census Commissioner, New Delhi Retrieved from
India aims to make agroforestry a major land use by intensifying its https://censusindia.gov.in/2011census/population_enumeration.html.
total agroforestry area to 53 M ha by 2050 through the restoration of fal- CEPF: Critical Ecosystem Partnership Fund, 2005. Ecosystem profile: North East-
lows, cultivable fallows, pastures, groves, and rehabilitation of degraded ern Himalayas Region Retrieved from https://www.cepf.net/sites/default/files/
final.ehimalayas.ep.pdf.
soils. This study integrates the relevant attributes of climate, topogra- Chavan, S.B., Keerthika, A., Dhyani, S.K., Handa, A.K., Newaj, R., Rajarajan, K., 2015.
phy, soil, ecological and socio-economy and presents a viable method National agroforestry policy in India: a low hanging fruit. Curr. Sci. 108 (10), 1826–
to assess the suitability of such land available in the EIHR. Given the 1834. doi:10.18520/cs/v108/i10/1826-1834.
Collins, M., Steiner, F., Rushman, M., 2001. Land-use suitability analysis in the United
enormous potential for agroforestry development in the region as re- States: historical development and promising technological achievements. Environ.
ported in this study, agroforestry can be a viable solution for addressing Manage. 28 (5), 611–621. doi:10.1007/s002670010247.
land degradation and climate change mitigation and adaptation while Dagar, J.C., Tewari, V.P., 2017. Evolution of Agroforestry as a Modern Science.
Springer Nature, Singapore Eds: Dagar, J.C. and Tewari, V.P. Agroforestry
ensuring farmers’ social, ecological, and economic sustainability in the
doi:10.1007/978-981-10-7650-3_2.
EIHR. Das, M., Nath, P.C., Sileshi, G.W., Pandey, R., Nath, A.J., Das, A.K., 2021. Biomass mod-
Land-use planners can use a multi-criteria approach amalgamating els for estimating carbon storage in Areca palm plantations. Environ. Sustain. Indic.
2021-10, 100115.
the biophysical, economic, social, political, and environmental issues
Defense Mapping Agency (DMA), 1992. Digital Chart of the World. Defense Mapping
for developing the agroforestry suitability model. Moreover, suitability Agency, Fairfax, Virginia.
evaluation can be considered for site-specific land evaluation based on Dhyani, S.K., Tewari, R.K., Tiwari, R., Singh, R., Srivas-
the relevant factors at a local scale. Besides, the site-specific local plants tava, S.R., 2013. Production and Guidance Retrieved from
http://cafri.res.in/html/CAFRI-AR/AR(2013-14)_English.pdf.
and crop species may be included for model building at a relatively fine Dhyani, S.K., 2014. National agroforestry policy 2014 and the need for area estimation
spatial scale to obtain more precise results. under agroforestry. CurrSci 107 (1), 9–10.

20
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Didan, K. (2015). MOD13A3 MODIS/Terra vegetation Indices Monthly L3 Global 1 km Laskar, S.Y., Sileshi, G.W., Pathak, K., Debnath, N., Nath, A.J., Laskar, K.Y., Singnar, P.,
SIN Grid V006 [Data set]. NASA EOSDIS Land Processes DAAC. Accessed 2020-07-27 Das, A.K., 2020. Variations in soil organic carbon content with chronosequence,
from 10.5067/MODIS/MOD13A3.006 soil depth and aggregate size under shifting cultivation. Sci. Total Environ.
Dragan, P.;., Zeljko, S.;., Sinisa, S., 2018. A new model for determining weight coefficient doi:10.1016/j.scitotenv.2020.143114.
of criteria in MCDM models: full consistency method (FUCOM). Symmetry (Basel) 10 Lopresti, M.F., Di Bella, C.M., Degioanni, A.J., 2015. Relationship between MODIS-NDVI
(9), 393. data and wheat yield: a case study in Northern Buenos Aires province. Argent Inf
Earth Resources Observation and Science Center/U.S. Geological Survey/U.S. Depart- Process Agric 2, 73–84. doi:10.1016/j.inpa.2015.06.001.
ment of the Interior, 1997. USGS 30 ARC-second Global Elevation Data, GTOPO30, Malczewski, J., 1999. GIS and Multicriteria Decision Analysis. JohnWiley& Sons, New
10.5065/A1Z4-EE71. Research Data Archive at the National Centre for Atmospheric York, p. 408.
Research, Computational and Information Systems Laboratory, Boulder, Colo Ac- Malczewski, J., 2004. GIS-based land-use suitability analysis: a critical overview. Prog
cessed† 11/10/2020. Plann 62 (1), 3–65.
Effat, H.A., Hassan, O.A., 2013. Designing and evaluation of three alternatives high- Mandal, D., Sharda, V.N., 2013. Appraisal of soil erosion risk in the Eastern Hi-
way routes using the analytical hierarchy process and the least-cost path analy- malayan Region of India for soil conservation planning. Land Degrad. Dev. 24,
sis, application in Sinai Peninsula. Egypt. Egypt J Remote Sens Space Sci 16 (2), 430–437.
141–151. Mbow, C., Van Noordwijk, M., Prabhu, R., Simons, T., 2014. Knowledge gaps and re-
Elsheikh, R., Mohamed Shariff, A.R.B., Amiri, F., Ahmad, N.B., Balasundram, S.K., search needs concerning agroforestry’s contribution to sustainable development goals
Soom, M.A.M., 2013. Agriculture Land Suitability Evaluator (ALSE): a decision and in Africa. CurrOpin Environ Sustain 6, 162–170.
planning support tool for tropical and subtropical crops. Comput. Elect. Agr. 93, 98– MEA: Millennium Ecosystem Assessment, 2005. Ecosystems and Human well-being: Bio-
110. doi:10.1016/j.compag.2013.02.003. diversity Synthesis. World Resources Institute, Washington, DC.
FAO (2008) Climate change and food security: A framework document. Report of the FAO Mishra, G., Giri, K., Pandey, S., 2018. Role of Alnus nepalensis in restoring soil fertil-
Interdepartmental Working Group on Climate Change. Rome. ity: a Case Study in Mokokchung. Nagaland. Natl. Acad. Sci. Lett. 41, 265–268.
FAO, 2017. Livestock Solutions For Climate Change Available at doi:10.1007/s40009-018-0668-4.
http://www.fao.org/publications/card/en/c/I8098EN/. Nair, P.K.R., 2011. Carbon Sequestration Studies in Agroforestry systems: a reality-check.
FAO/IIASA/ISRIC/ISS-CAS/JRC, 2012. Harmonized World Soil Database (Version 1.2). Agroforestry systems doi:10.1007/s10457-011-9434z.
FAO and IIASA, Rome, Italy and Laxenburg, Austria. Nath, A.J., Brahma, B., Sileshi, G.W., Das, A.K., 2018. Impact of land use changes
FAO: Food and Agriculture Organization of the United Nations, 2007. Land Evaluation: on the storage of soil organic carbon in active and recalcitrant pools in
Towards a Revised Framework. FAO, Rome, Italy. a humid tropical region of India. Sci. of the Tot. Environ. 624, 908–917.
Farzanpour, H., Shiri, J., Sadraddini, A.A., Trajkovic, S., 2019. Global comparison of 20 doi:10.1016/j.scitotenv.2017.12.199.
reference evapotranspiration equations in a semiarid region of Iran. Hydrol. Res. 50 Nath, A.J., Tiwari, B.K., Sileshi, G.W., et al., 2019. Allometric models for estimation of
(1), 282–300. doi:10.2166/nh.2018.174. forest biomass in North East India. Forests 10, 103. doi:10.3390/f10020103.
Fischer, G., Nachtergaele, F., Prieler, S., van Velthuizen, H.T., Verelst, L., Wiberg, D., Nath, A.J., Sahoo, U.K., Giri, K., Sileshi, G.W., Das, A.K., 2020. Incentivizing hill farmers
2008. Global Agro-Ecological Zones Assessment For Agriculture (GAEZ 2008). for promoting agroforestry as an alternative to shifting cultivation in Northeast India.
IIASA, Laxenburg, Austria and FAO, Rome, Italy https://www.fao.org/soils-portal/ Agroforestry for Degraded Landscapes In: Dagar, J.C., et al. (Eds.). Springer Nature
soil-survey/soil-maps-and-databases/harmonized-world-soil-database-v12/soil- Singapore Pte Ltd doi:10.1007/978-981-15-4136-0_14.
qualities-data/en/. Nath, A.J., Sileshi, G.W., Laskar, S.Y., Pathak, K., Reang, D., Nath, A., Das, A.K., 2021.
Fischer T., Beyerlee D., Edmeades G. (2014) Crop Yields and Global Food Security. Quantifying carbon stocks and sequestration potential in agroforestry systems under
Will Yield Increase Continue to Feed the World? ACIAR Monograph No. 158 Aus- divergent management scenarios relevant to India’s Nationally Determined Contribu-
tralian Centre for International Agricultural Research, Canberra, 2014 ISBN: 978-1- tion. J Clean Prod doi:10.1016/j.jclepro.2020.124831.
925133059, Xxii + 634 pp. NRCAF: National Research Centre for Agroforestry, 2013. The Vision-2050. Ministry of
Gebre, A.B., 2016. Potential Effects of Agroforestry Practices on Climate Change Mitiga- Agriculture, New Delhi.
tion and Adaptation Strategies: a Review. J. Nat. Sci. Res. 6 (15), 83–89 n. Pandey, R., Aretano, R., Gupta, A.K., Meena, D., Kumar, B., Alatalo, J.M., 2017.
Gomes, A.C.C., Bernardo, N., Alcantara, E., 2017. Accessing the southeastern Brazil 2014 Agroecology as a Climate Change Adaptation Strategy for Smallholders of Tehri-
drought severity on the vegetation health by satellite image. Nat. Hazards 89, 1401. Garhwal in the Indian Himalayan Region. Small-scale Forestry 16 (1), 53–63.
doi:10.1007/s11069-017-3029-6. doi:10.1007/s11842-016-9342-1.
Hancock, A.M., Witonsky, D.B., Alkorta-Aranburu, G., Beall, C.M., Gebremedhin, A., Suk- Press Information Bureau, Government of India. ( 2019 , September 9). India Will Re-
ernik, R., Utermann, G., Pritchard, J.K., Coop, G., Di Rienzo, A., 2011. Adapta- store 26 Million Hectares of Degraded Land by 2030 [Press release]. Retrieved from
tions to climate-mediated selective pressures in humans. PLos Genet. 7, e1001375. https://pib.gov.in/Pressreleaseshare.aspx?PRID=1584542.
doi:10.1371/journal.pgen.1001375. Purevtseren, M., Myagmarsuren, B., Brandt, S.A., 2017. Implications of a spatial multicri-
Hernandez, R.R., Debenport, S.J., Leewis, M.-.C.C.E., Ndoye, F., Nkenmogne, K.I.E., teria decision analysis for urban development in Ulaanbaatar, Mongolia. Mathemati-
Soumare, A., Thuita, M., Gueye, M., Miambi, E., Chapuis-Lardy, L., Diedhiou, I., cal Problems Engineering doi:10.1155/2017/2819795.
Dick, R.P., 2015. The native shrub, Piliostigma reticulatum, as an ecological “re- Ramos, N.C., Gastauer, M., Cordeiro, A.A.C., 2015. Environmental filtering of
source island” for mango trees in the Sahel. Agric. Ecosyst. Environ. 204, 51–61. agroforestry systems reduce the risk of biological invasion. Agrofor. Syst. 89,
doi:10.1016/j.agee.2015.02.009. 279–289.
Herzberg, R., Pham, T.G., Kappas, M., Wyss, D., Tran, C.T.M., 2019. Multi-Criteria decision Rana, S.K., Rana, H.K., Shrestha, K.K., Sujakhu, S., Ranjitkar, S., 2018. Determining bio-
analysis for the land evaluation of potential agricultural land use types in a hilly area climatic space of Himalayan alder for agroforestry systems in Nepal. Plant Div 40 (1),
of Central Vietnam. Land (Basel) 8 (6), 90. doi:10.3390/land8060090. 1–18.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A., 2015. Very high resolution Reisner, Y., De Filippi, R., Herzog, F., Palma, J., 2007. Target regions for silvoarable agro-
interpolated climate surfaces for global land areas. Int. J. Climatol. 25, 1965–1978. forestry in Europe. Ecol. Eng. 29 (4), 401–418.
Hopkins, L.D., 1977. Methods for generating land suitability maps: a comparative evalu- Roslee, R., Mickey, A.C., Simon, N., Norhisham, M.N., 2017. Landslide Susceptibility
ation. J. Am. Inst. Plann. 43 (4), 386–400. Analysis (LSA) Using Weighted OverlayMethod (Wom) Along The Genting Sem-
IFSR: India State of Forest Report (2019) Forest Survey of India, Ministry of Environment, pah To Bentong Highway. Pahang. Malaysian Journal Geosciences 1 (2), 13–19.
Forest and Climate Change, Government of India. doi:10.26480/mjg.02.2017.13.19.
IPCC, 2019: Climate Change and Land: an IPCC special report on climate change, deser- Saaty, T.L., Vargas, L.G., 2013. Sensitivity analysis in the analytic hierarchy process. In:
tification, land degradation, sustainable land management, food security, and green- Decision Making with the Analytic Network Process, 2013. Springer, New York, NY,
house gas fluxes in terrestrial ecosystems [P.R. Shukla, J. Skea, E. Calvo Buendia, USA, pp. 345–360.
V. Masson-Delmotte, H.-O. Pörtner, D. C. Roberts, P. Zhai, R. Slade, S. Connors, Sahoo, U.K., 2009. Traditional home gardens and livelihood security in North-East India.
R. van Diemen, M. Ferrat, E. Haughey, S. Luz, S. Neogi, M. Pathak, J. Petzold, J. Journal of Food. Agriculture Environment 7 (2), 665–670 200.
Portugal Pereira, P. Vyas, E. Huntley, K. Kissick, M. Belkacemi, J. Malley, (eds.)]. Sarkar, A., Ghosh, A., Banik, P., 2014. Multi-criteria land evaluation for suitability analysis
In press. of wheat: a case study of a watershed in the eastern plateau region, India. Geo-spatial
ISCGM: International Steering Committee for Global Mapping/Survey of India, 2016. Information Science 17 (2), 119–128. doi:10.1080/10095020.2013.774106.
Roads and Drainage, India. International Steering Committee for Global Mapping Sharma, E., Chettri, N., Tse-ring, K., Shrestha, A.B., Fang, J., Mool, P., Eriksson, M., 2009.
[Shapefile]Retrieved from https://maps.princeton.edu/catalog/stanford. Climate Change Impacts and Vulnerability in the Eastern Himalayas. ICIMOD, Kath-
IUSS (2014) World Reference Base for Soil Resources 2014. International Soil Classifi- mandu.
cation System for Naming Soils and Creating Legends for Soil Maps. World Soil Re- Sharma, P.D., 2004. Managing natural resources in the Indian Himalayas. J. Indian Soc.
sources Reports No. 106. FAO, Rome. Soil Science 52, 314–331.
Jain, S.K., Kumar, V., Saharia, M., 2012. Analysis of rainfall and temperature trends in Sileshi, G.W., Mafongoya, P.L., Akinnifesi, F.K., et al., 2014. Agroforestry: fertilizer
northeast India. Int. J. Climatol. doi:10.1002/joc.3483. trees. In: Encyclopaedia of Agriculture and Food Systems, 1. Elsevier, San Diego,
Jain, A., 2014. When did India–Asia collide and make the Himalaya? Curr. Sci. 106 (2), pp. 222–234.
254–266. Retrieved March 11, 2021, from http://www.jstor.org/stable/24099806. Singha, C., Swain, K.C., 2016. Land suitability evaluation criteria for agricultural crop
Jhajharia, D., Shrivastava, S.K., Sarkar, D., Sarkar, S., 2009. Temporal characteristics of selection: a review. Agricultural Reviews 2016 (37), 125–132.
pan evaporation trends under the humid conditions of northeast India. Agric. For. Szott, L.T., Fernandes, E.C.M., Sanchez, P.A., 1991. Soil-plant interactions in agroforestry
Meteorol. 149, 763–770. systems. For. Ecol. Manage. 45 (1), 127–152. doi:10.1016/0378-1127(91)90212-E.
Joss, B.N., Hall, R.J., Sidders, D.M., Keddy, T.J., 2008. Fuzzy-logic modeling of land suit- Torabian, S., Soffianian, A., Fakheran, S., et al., 2018. Habitat suitability mapping for sand
ability for hybrid poplar across the Prairie Provinces of Canada. Environ. Monit. As- cat (Felis margarita) in Central Iran using remote sensing techniques. Spat. Inf. Res.
sess. 141, 79–96. doi:10.1007/s10661-007- 9880-2. 26 (2018), 11–20. doi:10.1007/s41324-017-0152-0.

21
A.J. Nath, R. Kumar, N.B. Devi et al. Environmental Challenges 4 (2021) 100199

Tzioutzios, C., Kastridis, A., 2020. Multi-Criteria Evaluation (MCE) Method for the Man- Zolekar, R.B., Bhagat, V.S., 2015. Multi-criteria land suitability analysis for agriculture in
agement of Woodland Plantations in Floodplain Areas. ISPRS Int. J. Geo-Inf. 2020 (9), hilly zone: remote sensing and GIS approach. Comput. Electron. Agric. 2015 (118),
725. doi:10.3390/ijgi9120725. 300–321.
Walter, I.A., Allen, R.G., Elliott, R., Jensen, M.E., Itenfisu, D., Mecham, B., et al., 2000. Zomer, R.J., Trabucco, A., Bossio, D.A., van Straaten, O., Verchot, L.V., 2008. Climate
ASCE’s standardized reference evapotranspiration equation. In: Proc. 4th Nat’l. Irrig. Change Mitigation: a Spatial Analysis of Global Land Suitability for Clean Develop-
Symp., ASAE. Phoenix, AZ. ment Mechanism Afforestation and Reforestation. Agric. Ecosystems and Envir. 126,
Yen, B.T., Visser, S.M., Hoanh, C.T., Stroosnijder, L., 2013. Constraints on agricultural 67–80.
production in the northern uplands of Vietnam. Mount. Res. Develop. 33, 404–415.
doi:10.1659/MRD-JOURNAL-D-13-00015.1.

22