







































 

Ecology, Economy and Society–the INSEE Journal 1 (2): 31–65, July 2018 

 
RESEARCH PAPER 

 

Modelling the Economics of Grassland Degradation 
in Banni, India, using System Dynamics 
 

Mihir Mathur  and Kabir Sharma  
 
Abstract: This is a study of the interactions between the ecology and economy of 
the Banni grassland, located in the district of Kutch, Gujarat, India. The study 
focuses on modelling the economic impact of grassland degradation in the Banni 
from 1992–2015 and simulates future scenarios up to 2030 using system dynamics. 
The specific sectors being modelled are the area spread of the invasive species 
Prosopis juliflora, palatable grass, the populations of livestock as well as the livestock 
and charcoal incomes of Banni. An economic valuation is done by discounting the 
future earnings of the pastoral (milk, livestock sale, dung manure) and charcoal 
economy under two scenarios 1) Base case (Business as Usual), i.e. keeping current 
policies constant and 2) P. juliflora removal policy (PRP) i.e. where a decision is 
implemented to remove P. juliflora from Banni. Under the BAU scenario, modelling 
results indicate that the Banni grassland is headed for severe fodder scarcity due to 
the shrinking area under grassland. Under the PRP scenario, Banni is able to revive 
its grasslands and increase the present value of future earnings (up till 2030) by 62 
per cent. A delay of five years in the decision to remove P. juliflora results in a 28 
per cent reduction in earnings indicating the policy’s time sensitivity. The model 
serves as a test bed for generating what-if scenarios of the Banni grassland. 

Keywords: Grasslands; livestock; system dynamics; economics; land degradation 

 
 Founder Director, DESTA Research LLP, New Delhi; mihir09mathur@gmail.com.  
 Founder Director, DESTA Research LLP, New Delhi; sharmakabir@gmail.com 

This study was carried out in year 2016, when the authors were working at The Energy and 
Resources Institute (TERI), New Delhi, as a part of the project ‘Economics of 
Desertification, Land Degradation and Drought in India’ funded by Ministry of 
Environment, Forest and Climate Change, Government of India. Then, Mihir was Associate 
Fellow, Center for Global Environment Research and Kabir was Research Associate, 
Modelling and Scenario Building.  

Copyright © Mathur and Sharma 2018. Released under Creative Commons Attribution-
NonCommercial 4.0 International licence (CC BY-NC 4.0) by the author.  

Published by Indian Society for Ecological Economics (INSEE), c/o Institute of Economic 
Growth, University Enclave, North Campus, Delhi 110007.  

ISSN: 2581-6152 (print); 2581-6101 (web). 

DOI: https://doi.org/10.37773/ees.v1i2.34  

https://doi.org/10.37773/ees.v1i2.34


Ecology, Economy and Society–the INSEE Journal [32] 

 

1. INTRODUCTION 

Changes in ecosystems, climate, and peoples’ livelihoods are all 
interdependent. The way in which biological resources are converted into 
goods and services, as inputs and outputs of the economy, contributes to 
shifts in local livelihoods and to climate change (Perrings 2010, 3). The loss 
of biotic species, reduced biomass, and, thereby, reduced carbon 
sequestration contributes to climate change and impacts the livelihoods of 
people dependent on such resources, especially grasslands and forests. 
Interventions designed at the policy or community level to induce changes 
in ecosystems often generate economic and social feedback by changing the 
livelihoods and lifestyles of resource-dependent communities after a long 
time delay. This is more so in the case of large-scale changes made to 
convert forests to croplands or grasslands to forests. 

One such example of this coupling between humans, the economy, and 
ecosystems is seen in tropical semi-arid grasslands, wherein the livelihoods 
of dwelling communities are deeply interlinked with ecological dynamics 
(Reid et al. 2005, 56). In India, in recent times, there has been the tendency 
of policymakers to misconceive grasslands for degraded forests or 
wastelands; planting up grasslands or converting these to industry deeply 
impacts the social-ecological dynamics of these regions (Vanak et al. 2017). 
In the Banni grassland in Kutch, Gujarat, seeding in the 1960s of Prosopis 
juliflora (P. juliflora), an alien invasive woody species, has over time greatly 
hampered grass production and the traditional livelihoods of nomadic 
pastoralists who have lived there for generations (Bharwada and Mahajan 
2012, 76–77). In the context of climate change it is even more important to 
understand the dynamics between the socio-economics and ecology of such 
systems—to effectively address conflicting climate goals, such as increasing 
carbon sinks through afforestation while conserving pastoral livelihoods 
and grassland biodiversity, and to reach informed decisions on their 
management (Xu et al. 2014). In such a scenario, a systemic 
understanding—seeing the social-economic and ecological dynamics 
together as one system—is critical. This warrants the use of systems 
thinking and system dynamics simulation modelling methods. 

Through this paper the authors demonstrate the potential of system 
dynamics for modelling complex social, ecological, and economic 
interlinkages in coupled human-natural systems such as Banni. The aim of 
this study is to estimate the economic cost of grassland degradation in 
Banni over a 15-year period (2015–2030). It achieves this through 
modelling the dynamics and feedbacks between the grassland, P. juliflora 
spread, livestock populations, and the economics of the Banni grassland. 



[33] Mihir Mathur and Kabir Sharma 

 

The model highlights the interdependencies between different sectors and 
between variables within each sector. The model helps in developing a 
deeper understanding of the complexities of Banni and serves as a tool for 
generating what-if scenarios for various policies. The study highlights the 
need for further research on the ecological and economic parameters of 
Banni, and presents a case for the development of a decision support tool 
for developing sustainable management plans of the Banni grassland. 

1.1 Banni Grassland 

The Banni grassland, once known as Asia’s best tropical grassland 
(Bharwada and Mahajan 2012, 6), spans an area of approximately 2,500 sq 
km in the district of Kutch on the northern border of Bhuj block (Koladiya 
et al. 2016, 9). The grassland has been degrading; productivity dropped from 
4,000 kg per hectare in the 1960s to 620 kg per hectare in 1999 (Bharwada 
and Mahajan 2012, 76). While the evidence is still inconclusive on whether 
the dominant cause is increasing salinity or the spread of the invasive 
species P. juliflora, the most cited reason by the pastoralists (Maldharis) is 
the spread of P. juliflora. Livestock rearing is the primary occupation of the 
people, and grassland degradation poses a serious problem for sustaining 
the pastoral economy. 

The Banni grassland is divided into Ugamani Banni or East Banni; Vachali 
Banni or Central Banni; and Aathamani Banni or Jat Patti, or West Banni 
(Bharwada and Mahajan 2012, 12). Its geographic area has been estimated, 
variously, at 1,800 sq km to 3,800 sq km (Bharwada and Mahajan 2012, 
125); this study uses the recent estimate of 2,500 sq km (Koladiya et al. 
2016, 20). 

The dominant community is the livestock-breeding Maldharis. Many 
communities among them—Raysipotra, Halepotra, Pirpotra, Hingorja, 
Sumra, Mutva, Node, etc.—migrated several generations ago from Sindh, 
Marwar, and Baluchistan. The other community is the Meghwals, and their 
main occupations have been leather tanning and shoemaking and making 
artefacts from leather (Bharwada and Mahajan 2012, 14). 

1.2. The Banni Grassland—A Brief History 

The mainstay of the Maldhari community of Banni has been livestock 
breeding. Before independence the Maldhari communities lived a nomadic 
life and often wandered into what is now Pakistan for grazing their 
livestock (Bharwada and Mahajan 2012, 51). 

Banni is largely flatland. As it is often flooded in the monsoon, especially 
west Banni, even spurring intra-Banni migration, it is sometimes referred to 



Ecology, Economy and Society–the INSEE Journal [34] 

 

as a seasonal wetland (Mehta et al. 2014, 18). The government forest 
department reports about 254 small and large wetlands (Bharwada and 
Mahajan 2012, 41). In May 1955, Banni was declared a protected forest 
under the Indian Forest Act, 1927 (Mehta et al. 2014, 20). Climatically, it is 
an arid or semi-arid zone; annually, it experiences around 300–353 mm of 
rainfall on average (Bharwada and Mahajan 2012, 114; Geevan et al. 2003, 
21). 

1.2.1. Prosopis juliflora 

P. juliflora is a species native to South America, the Caribbean, and Mexico. 
The forest department first introduced it over an area of 31,550 hectares 
along the border between Banni and the Great Rann of Kutch in 1961 to 
control the ingression of the Rann (salt marsh) into other land (Bharwada 
and Mahajan 2012, 83). But the invasive nature of P. juliflora and its spread 
over the past 55 years has led to the loss of native vegetation, including 
grassland; today, the Maldharis cite it as one of the dominant causes of 
grassland degradation.  

In the summer, when grasses are in short supply, the livestock feed on P. 
juliflora pods. Seeds, rejected in their fecal matter, receive both manure and 
moisture in this way, and quickly take root and germinate (Kumar 2015, 47). 
This helps P. juliflora spread further. The open grazing system accelerates 
the process (Bharwada and Mahajan 2012, 83). P. juliflora apportions 
resources and also has allelopathic properties. Due to these reasons its 
spread has reduced the area under indigenous plants and grassland in Banni 
(Bharwada and Mahajan 2012, 81). The loss in grassland productivity has 
raised the amount of fodder that the Maldharis purchase and, thus, 
livestock expenses for the economy. Pastoralists opine that the grasslands 
would recover if P. juliflora—locally Gando Baval, or ‘mad’ Acacia—were to 
be removed.  

1.2.2. Dairy 

The Banni buffalo and Kankrej cattle are the dominant large livestock 
species in Banni. Traditionally, Banni was not a dairy-farming economy; its 
pastoralists were livestock breeders and involved in the trade of the Banni 
buffalo and Kankrej cattle and bullocks. Only after dairies began collecting 
milk in 2009–2010 did the pastoralists start selling milk in large quantities. 
In 2010, Banni’s buffalo was recognized by the National Bureau of Animal 
Genetic Resources as a distinct breed; and this increased its value and 
spurred its rearing (Bharwada and Mahajan 2012, 98). Consuming the pods 
of P. juliflora is fatal for Kankrej cattle; therefore, their population has been 
decreasing (Bharwada and Mahajan 2012, 87) although, in recent years, the 



[35] Mihir Mathur and Kabir Sharma 

 

Maldharis have observed that the Kankrej have been adapting to P. juliflora 
and surviving in areas that are dense with it. The Banni buffalo is 
unaffected by P. juliflora, and their numbers have grown, which has helped 
the dairy business since buffalo milk (which has greater fat content) fetches 
higher prices.  

1.2.3. Charcoal Making 

The Maldharis practice charcoal making to earn additional income and 
harvest the wood of P. juliflora for this purpose. It is illegal to cut P. juliflora 
in the grassland, as it is classified as a protected reserve forest. When this 
ban was lifted between 2004 and 2008, the Maldharis took to removing P. 
juliflora trees from the roots for making charcoal, which led to a huge 
reduction in the area under P. juliflora, as uprooting the trees freed up land 
and allowed grasses to grow. It is hard to calculate the exact increase in 
charcoal production in this period, but it is estimated at as high as 10 times 
(Bharwada and Mahajan 2012, 100). In 2008, this ban was re-imposed—
some suggest because indigenous trees were also being harvested for 
charcoal; others suggest that the charcoal traders’ cartel influenced it 
because they were unable to exercise control over production and supply of 
charcoal (Bharwada and Mahajan 2012, 101). The ban persists, but charcoal 
making continues. 

1.3. Research Objective 

The main research objective of this study is to model the economics of 
grassland degradation in Banni due to the invasion of P. juliflora. It focuses 
on the issue of grassland degradation, its key drivers, and what impact the 
removal of P. juliflora would have as a solution for halting grassland 
degradation. The study’s objective does not include modelling climate 
change impacts on these future scenarios, although a sensitivity analysis of 
rainfall variation has been carried out. 

 

2. RESEARCH METHODOLOGY: SYSTEM DYNAMICS 

Banni’s ecological and economic system is interdependent and highly 
dynamic. Thus, the research method uses system dynamics modelling to 
develop a base case and policy scenarios of Banni. System dynamics is a 
modelling approach aimed to understand the nonlinear behaviour of 
complex systems over time using stocks and flows, internal feedback loops, 
and time delays (MIT 1997). Beginning at the Massachusetts Institute of 
Technology (MIT) in the 1950s and 1960s (Forrester 1958; Urban 
Dynamics 1969), system dynamics as a method unveils the counterintuitive 



Ecology, Economy and Society–the INSEE Journal [36] 

 

nature of complex systems and uncovers relationships between variables 
that are responsible for the behaviour of a system over time. System 
dynamics offers the potential for generalization since it provides a more 
transparent and consistent description of the underlying causal processes of 
the system (Gallati and Wiesmann 2011, 8). This is especially useful for 
developing policy insights and systemic decision-making. 

System dynamics and systems thinking have been used to study grassland-
livestock systems worldwide. Some have been used to understand social-
ecological dynamics in the agro-pastoral Sahel (Rasmussen et al. 2014; 
Sendzimir et al. 2011; Prado et al. 2014). Others have been used to model 
natural resource-based poverty traps in the context of small-holder farmers 
in highland Kenya (Stephens et al. 2011). It has been used in biodiversity 
conservation in semi-arid savannas in Central and Southern Africa (Perrings 
and Walker 1999), and in understanding the dynamics of urbanization and 
environmental policy on the future availability of grazing resources in the 
Mongolian plateau (Allington et al. 2017). In India, system dynamics has 
been used to understand the dynamics of social-ecological systems, most 
pertinently in studying the effect of P. juliflora on diverse agents in a national 
park (Dayal 2007), and in studying the grassland systems of Kutch, 
including Banni, from an ecological-economic perspective (Geevan et al. 
2003). 

Various aspects of the Banni grassland have been studied and documented 
(Koladiya et al. 2016; Jagruti et al. 2017; Shah et al. 2010), no study attempts 
a bottom-up ecological-economic model, and only a few studies integrate 
the ecology and socio-economics. Two such integrated studies (Bharwada 
and Mahajan 2012; Geevan et al. 2003) have been used as base papers for 
this study. This paper aims to partially fill this gap in the process of meeting 
its aim of studying the economics of grassland degradation. 

We present a description of our system dynamics model, built in a bottom-
up manner for the specific case of the Banni grasslands, taking inputs from 
academic experts, the Maldhari community, as well as NGOs working in 
Banni. The model considers Banni as one system and does not model the 
spatial differences within it. This model comprises livestock (Banni buffalo 
and Kankrej cattle); grassland and P. juliflora; and the pastoral and charcoal 
economy. We model the impacts of drivers of livestock growth and P. 
juliflora growth, their impact on the local environment, and the consequent 
multiple feedbacks that could impact the future of these sectors. The model 
runs from 1992 to 2030, recreating the major dynamics seen between 1992 
and 2014 and simulating future scenarios up to 2030 under a base case; 
policy implementation of P. juliflora removal; and a five-year-delay in the 



[37] Mihir Mathur and Kabir Sharma 

 

implementation of the policy of removing P. juliflora. The paper presents the 
higher order causal loop diagram of the simulation model, its model 
description, key feedbacks and assumptions, and the simulation results and 
insights generated, respectively.  

The higher order causal loop diagram (Figure 1) shows higher order 
linkages of our system dynamics model. There are two balancing loops 
(livestock–fodder–livestock, livestock–P. juliflora–grassland–livestock) and 
two reinforcing loops (livestock–total income–profitability–livestock and 
livestock–P. juliflora–total income–profitability–livestock). 

Figure 1: Higher Order Causal Loop Diagram of the Simulation Model 

 

 

 

 

 

 

 

 

 

 

 

 

 

The first balancing process (balancing loop B1) is concerned with the 
spread of P. juliflora. The growth in the area under P. juliflora is aided by the 
presence of livestock, which act as vectors, carry the seeds, and help in their 
widespread dispersal. As the area under P. juliflora goes up, the area under 
grassland comes down—leading to a negative impact on livestock due to 
falling fodder availability. On the other hand, an increase in livestock 
numbers raises the fodder requirement. With limited grassland, fodder 
availability falls, the Maldharis purchase more feed and fodder, input costs 
rise, and the profit per livestock falls. If it falls below a threshold, it results 
in a stress sale of livestock for recovering losses, balancing out the livestock 



Ecology, Economy and Society–the INSEE Journal [38] 

 

numbers. As the livestock numbers drop, the fodder requirement falls, and 
livestock profitability rises. These dynamics form balancing loop B2. More 
livestock would yield more milk, thereby increasing income and forming 
reinforcing loop 1, R1. Higher prices for Banni buffalo milk constitute the 
key driver for the growth of profitability. 

The other reinforcing loop, R2, is concerned with the charcoal income 
earned from P. juliflora. The P. juliflora spread is accelerated by the presence 
of livestock; they act as vectors, and increase the area under P. juliflora 
charcoal-producing potential, and income, which makes further stocking of 
livestock possible. But the increase in the area under P. juliflora reduces the 
grassland area and fodder, as discussed in balancing loop B1. Interestingly, 
thus, livestock is a common driver for both the reinforcing and balancing 
loops. 

Currently, the reinforcing loops R1 and R2 are dominant, as Maldharis are 
earning enough from milk sales, livestock sale, dung, and charcoal to sustain 
their livestock, even with grassland degradation and the consequent fodder 
deficit. Good milk income and the presence of dairies in Banni provide 
Maldharis the economic incentive to retain and grow their livestock 
numbers. However, balancing loops B1 and B2, less impactful presently, 
could become dominant in the future, and greatly impact dynamics. 

Policy testing is done for a case of P. juliflora removal, which is shown as an 
external variable in Figure 1. The removal of P. juliflora could potentially 
reverse the current trend of grassland degradation and increase grassland 
area and fodder availability. 

2.1. Simulation Model and Sector Description 

The model comprises three interconnected sectors: livestock (buffalo and 
Kankrej cattle), P. juliflora and grassland, and the economy. Important 
parameter values are provided in the Annexure. 

The total area of Banni is taken as 2,500 sq km, i.e. 250,000 hectares 
(Koladiya et al. 2016, 20). Of this, 90 per cent is taken to be total possible 
productive land (includes grassland, P. juliflora-dominated area, and the area 
under other vegetation), while 10 per cent is taken to be wasteland where 
palatable grasses cannot grow (wasteland includes extreme saline land). In 
1992 (the base year), the area of land dominated by P. juliflora is taken to be 
41,180 ha (Koladiya et al. 2016, 20) while grassland area equals total 
productive area (less the area under P. juliflora). 

During princely rule, the Banni grasslands had a more structured, customary 
regime of governance and management, including certain grazing 



[39] Mihir Mathur and Kabir Sharma 

 

regulations (e.g. kinds of animals and the period of grazing) which were 
followed by the Maldharis (Geevan et al. 2003, 27). In those times, Banni 
was more like a managed common property resource. These governance 
structures have weakened since; and now different communities use Banni 
for various purposes (constructing resorts, experimenting with rainfed 
agriculture). Uncontrolled growth of buffalo herds also implies exploitative 
use of the grazing land. The grassland system has evolved into an open 
access resource, and it has been considered as such in the model. 

The dynamics between grassland and the area under P. juliflora is the key 
factor for changes in the grassland. The spread of P. juliflora is the main 
driver of land use change. As the area under P. juliflora grows, it invades the 
area under grassland. The normal spread rate of P. juliflora is taken to be 8.5 
per cent per year of the total area under P. juliflora (Vaibhav et al. 2012, 3). 
However, this spread rate is enhanced by the presence of livestock which 
act as vectors, the seeds being carried by them and the passage through the 
digestive tract facilitating quick germination (Geevan et al. 2003, 17; Kumar 
et al. 2015, 47). This has been modelled as a multiplier through a graphical 
function in our model—the impact increasing with increase in livestock 
population, and ultimately levelling off at an assumed maximum. The 
graphical function is given in Figure 2. The growth of P. juliflora is limited 
by the total land area available. 

In the model, it is assumed that Maldharis use the above-ground wood of P. 
juliflora for charcoal making, which does not reduce the area under P. 
juliflora. Only once before has the P. juliflora area came down when the ban 
on making charcoal from P. juliflora was lifted and people uprooted the trees 
entirely. This was between 2004 and 2008, and has been modelled using a 
time-based “if-then” function. The difference equation governing the 
change in area under P. juliflora in a time interval ∆T (annually in our case) is 
given below.  

TR
L

Ap
nEApAp −








−= ]1[  

Where, 

n = Normal annual spread rate of P. juliflora 
E= Multiplier on normal spread rate due to presence of livestock vectors 
(Graphical Function shown in Figure 2) 
Ap= Area under P. juliflora 
L= Total productive land area 
R= Clearing of area under P. juliflora through uprooting (during period of 
uplifting of the ban and under the P. juliflora removal policy) 



Ecology, Economy and Society–the INSEE Journal [40] 

 

Figure 2: Impact of Livestock on P. Juliflora 
Spread Rate 

 

The grassland biomass 
production in a year is 
calculated from the 
grassland area (total 
productive land less area 
occupied by P. juliflora) and 
the grassland productivity, 
the latter being a bell-
shaped function of the 
rainfall. Personal inter-
views revealed that the 
grassland productivity of 

Banni is high in a specific bandwidth of rainfall, and lower on both 
extremes (low and very high rainfall). 

Rainfall data for the grassland is taken from three different sources, because 
publicly available data (free of cost) for all required years was not found at 
any one source. The website of the India Meterological Department (IMD) 
gives weather data for Kutch district from 2012 to 2017 (IMD, n.d.). 
Bharwada and Mahajan (2012, 143) collate rainfall data for 1981 to 2010 
from two sources, Gujarat Institute of Desert Ecology (GUIDE) (up to 
1994) and the Irrigation Department and Rudramata Dam site 1994 
onwards (see pg. 143, appendix 9). The gap in data for year 2011 and 2012 
is filled in from Gavalli (2015, 5). Rainfall data at the Banni grassland’s 
spatial scale is not available for the model time period. For the business as 
usual (BAU) scenario, the rainfall pattern for 2015–30 is taken to be a 
repeat of 2001–15, to account for the cyclic variation that exists in Banni’s 
rainfall (Bharwada and Mahajan 2012, 143). However, realizing this to be 
insufficient, sensitivity runs by changing rainfall parameters for 2015–2030 
under three scenarios is done. Similar sensitivity runs are performed by 
changing the grass productivity parameter. Both are presented in the results 
section. 

Further, a parameter ‘fodder deficit’ is calculated, as the ratio between the 
fodder available in Banni in a particular year less the fodder requirement in 
that year divided by the fodder requirement. This ratio is important as it 
determines the input cost (feed and fodder purchased from outside Banni) 
for sustaining the livestock economy. As the deficit increases, the livestock 
input cost increases. This ratio also determines the migration of livestock 
from Banni in fodder deficit years, explained below. 

 

 



[41] Mihir Mathur and Kabir Sharma 

 

2.1.1. Livestock Dynamics 

This sector consists of populations of the two large ruminants: the Banni 
buffalo and Kankrej cattle. Small ruminants such as sheep and goats, 
though present in Banni are excluded due to their relatively smaller share of 
the total livestock (less than 10 per cent). For both the livestock (buffalo 
and cattle), modelling has been done by making ageing chains i.e. breaking 
down the populations into calves and adults, considering a maturation time 
and taking different death rates/retiring times for both stocks. Calves are 
born to a certain fraction of the adults every year. Some calves die before 
they transit into adults according to a calf death rate. There is also a retiring 
time for the adults after which they stop producing milk and giving birth to 
calves. 

To manage the frequent droughts in Banni, the Maldharis have adopted two 
dominant coping mechanisms. One is migrating out of Banni with their 
livestock for the dry period and the second is by increasing the sale of 
livestock in dry years. Based on discussions with Maldharis, it is estimated 
that if the fodder deficit crosses 30 per cent in a certain year, 30 per cent of 
the livestock leaves Banni, and if it crosses 50 per cent, 50 per cent of 
livestock leaves Banni. Also, the livestock that migrate outside accumulate 
in stocks of migrated livestock which come back when the deficit falls 
below 10 per cent. A maximum residing time of two years is given to the 
migrated stock of adult livestock after which the migrated stock 
permanently migrates out of the Banni periphery. The second coping 
mechanism is sale of livestock in dry years. A livestock sale multiplier is 
built using graphical function which depicts the impact of falling 
profitability on the flow of livestock (stress) sales. This sale multiplier 
depends on the net income per livestock. As the profit per livestock in a 
year becomes negative, the stress sale multiplier increases and later levels 
off. The values are provided in Figure 3. 

The Banni buffalo ageing chain is composed of two main stocks: Calves 
and Adults. The stock of buffalo calves has one inflow (births), two 
outflows (calf deaths, maturation to adult buffaloes) and one bi-flow (calf 
migration). The births are governed by a certain fraction of the adult 
buffaloes which give birth to a calf every year (approx. 50 per cent of the 
total adult stock). 50 per cent of the births are female and 50 per cent male. 
The model considers only females, as males are generally not reared. The 
fraction of buffalo calf death every year is taken as 20 per cent (after 
discussions with Maldharis). Maturation time from calf to adult is taken as 3 
years. The adult lifetime is taken as 20 years and sale rate of buffaloes is 
assumed at 1 per cent per year (based on interviews). The coupled 



Ecology, Economy and Society–the INSEE Journal [42] 

 

Figure 4: Impact of P. juliflora 
Density on Kankrej Death Rate 

 

difference equations governing the changes in stocks of buffalo calves and 
adults are given below for a time interval ∆T (annual in our case). 

TMC
Tm

C
dCfgBC 








−−−=  

TsPBMBrB
Tm

C
B 








−−−=  

Where, 

C= Stock of Calves 
B= Stock of Adult Buffaloes 
f= Fraction of adult buffaloes giving birth in a year 
g=Fraction of females in the births 
r = Retiring time of adult buffaloes 
d= Death rate of calves 
Tm= Maturation time 
s=Normal sale fraction per year 
M= Sale multiplier function from fodder deficit (Figure 3) 
P=Sale Multiplier from net income per livestock 
 

Figure 3: Impact of Net Income per livestock on Buffalo and Kankrej Sale 

The Kankrej ageing chain and 
equations are very similar to the 
buffalo ones, having birth fraction, 
lifetime, maturation time, fodder 
requirement etc. The main 
difference is that there exists a 
practice in Banni of purchasing 
Kankrej calves every year and as 
the Kankrej calves are very 
valuable, the stress sale function 
due to profitability (a function of 



[43] Mihir Mathur and Kabir Sharma 

 

livestock profitability, similar to buffaloes above) is of Kankrej calves and 
not adults. Another distinguishing feature is that the Kankrej cattle 
population is negatively affected by P. juliflora, as the cattle are unable to 
digest the pods and die on consuming them. This relationship is shown 
through a graphical function where the death multiplier increases due to 
increase in P. juliflora density. The values for two graphical functions 
discussed in this section are provided in Figure 3 and Figure 4. 

2.1.2. The Economy Dynamics 

This sector consists of livestock based income and charcoal based income. 
The model considers a uniform economic agent, the Maldharis, earning 
both from livestock based income and from charcoal based income. The 
milk income is composed of income from milk, dung, and livestock sale. 
The milk income is derived from the milk produced by the cows and the 
buffaloes, the livestock sale income from the fraction of cows/buffaloes 
sold in a year (which is a function of the profitability in a year, which is 
linked to rainfall, as described later in this section), and the income from 
dung sale from the existing number of livestock in a year. Forecasting 
future prices, at local level, has lot of uncertainty which would add to the 
complexity of carrying out an economic valuation of Banni grasslands. 
Hence, prices have been assumed to be constant at 2015 levels. There are 
costs of rearing livestock (largely of feed and fodder purchased from the 
external market) which were calculated based on interviews, and are 
factored into the calculation of net incomes. 

The charcoal annual income is derived from the annual charcoal 
production. Charcoal production in a year has been differentiated according 
to history. Before the ban on charcoal production was lifted (i.e. before 
2004): the charcoal production is taken as 2,400 sacks of 40 kg each per day 
for 240 days in a year (based on discussions with local NGO and personal 
interviews with Maldharis). During the time when the ban was lifted 
(between 2004 and 2008): the charcoal production is increased by 10 times 
as compared to before the ban (Bharwada and Mahajan 2012, 100), and 3) 
After the ban was again imposed (i.e. after 2008): the charcoal production is 
taken as 4,800 sacks of 40 kg each produced per day for 240 days in a year, 
the same as during the pre-ban period. A feedback function is created to 
increase the rate of production in the event of a fall in profits from 
livestock. The charcoal production numbers are difficult to collect from 
field owing to its fuzzy legal status. Thus, we believe the model numbers are 
conservative and underestimate the actual quantum of charcoal activity and 
income. 



Ecology, Economy and Society–the INSEE Journal [44] 

 

Figure 5: Impact of Fodder Deficit on 
Livestock input cost 

 

Summing the income from livestock and charcoal less the costs of livestock 
rearing and dividing by the total livestock population, a number for net total 
income per livestock is calculated. This number creates a feedback and 
governs the stress sales of adult buffaloes and Kankrej calves. As the net 
total income per livestock in a year goes negative, the stress sale multiplier 
increases, increasing the income from sales in the year, and reducing the 
buffaloes and cow numbers, both of which in the real world help the 
Maldharis to cope. 

The remittances, in the form of income from milk sales from migrated 
livestock, are considered in the model, and both the costs of rearing them 
and revenue from their milk sale are included. It is assumed that the cost of 
rearing the livestock remains the same even after migration, and thus 
migration is important for the Maldharis in order to be able to have access 
to feed and fodder, which falls in dry years within Banni’s geographic 
boundary. 

2.1.3. Key feedback relationships 

There are six cross-sectorial 
feedback loops which govern the 
dynamics of the model. 

1) Impact of fodder deficit on 
livestock input cost. As the 
fodder deficit increases so does 
the livestock input cost, 
reflecting the need to purchase 
fodder from outside. The 
relationship is given in Figure 5. 

2) Impact of net income per 
livestock on livestock stress sale rate. As the profit per livestock becomes 
negative, the stress sale of livestock goes up, reflected in an increase in 
stress sale fraction. This is depicted in Figure 3. 

3) Impact of fodder deficit on temporary livestock migration. If the 
fodder deficit lies between 30 per cent and 50 per cent in a certain year, 30 
per cent of the livestock leave Banni, and if it crosses 50 per cent, 50 per 
cent of livestock leave Banni. If fodder deficit is 10 per cent or lower, the 
livestock migrate back to Banni. 

4) Impact of livestock on P. juliflora spread rate. As the livestock 
population increases it leads to increase in the spread rate of area under P. 
juliflora. This is depicted in Figure 2. 



[45] Mihir Mathur and Kabir Sharma 

 

Figure 6: Impact of net income per 
livestock on Charcoal Production 

 

5)  Impact of P. juliflora density on Kankrej death rate. As the P. juliflora 
density (area under P. juliflora ÷ total productive area) increases it leads to 
an increase in Kankrej death rate. However, it has been observed by the 
Maldharis that Kankrej has adapted to survive in P. juliflora dense areas. 

Thus the death multiplier evens 
out at high levels of P. juliflora. 
This is discussed in Figure 4. 

6)  Impact of profit per livestock 
on charcoal production. As the 
profit per livestock becomes 
negative, charcoal production 
starts increasing to compensate for 
the losses. This is depicted in 
Figure 6. 

2.2. Model Assumptions 

1. Prices for milk, livestock, feed, charcoal, and dung manure are kept 
constant at 2015 levels. Forecasting future prices, at local level, has lot of 
uncertainty which would add to the complexity of carrying out an 
economic valuation of Banni grasslands. Hence, here it is assumed to be 
constant at 2015 prices. 

2. No limit on external supply of feed, fodder, and water. Today, an 
external supply of feed and fodder is an integral part of Banni’s economy 
and is assumed to be available for purchase at a cost. Water is available in 
Banni through pipelines coming in from outside the Banni boundary, and 
is assumed to remain sufficient for the duration of model runs. 

3. Exclusion of small ruminants (e.g. sheep, goat etc.). Buffalo and cattle 
constitute most of the Banni livestock. In 2011 their share was around 92 
per cent of the total livestock (Bharwada and Mahajan 2012, 88). Hence, 
considering the small proportion of small ruminants they are excluded 
from the study. 

4. Ecological impacts of P. juliflora removal: The model does not capture 
the ecological impacts of removing P. juliflora from Banni grassland as the 
scientific and practical knowledge on it is found to be inconclusive. 



Ecology, Economy and Society–the INSEE Journal [46] 

 

Figure 7: Base Case Livestock Population: 1992–2030 

 
 

3. RESULTS 

3.1. Base Run: Business as usual scenario 

The business as usual scenario i.e. base run simulation, indicates that the 
total livestock population (sum of buffaloes and Kankrej) in Banni would 
fall between 2016 and 2030 (Figure 7). The primary reason for this is 
increasing livestock stress sales and out migration due to reducing area 
under grassland and consequently decline in fodder availability. Two 
consecutive years of poor rainfall (2019–2020) are the reasons for the steep 
fall in livestock numbers in year 2020 similar to what was observed in year 
2004. Thus, livestock variability could be higher in periods of fodder 
scarcity. 

The shrinking area under grassland, due to P. juliflora spread, is a cause of 
concern for Banni (Figure 8). If current conditions persist then by year 2030 
the model shows that the area under grassland will reduce to approx. 27,000 
hectares from 92,000 hectares in 2015, a reduction of around 70 per cent. 
The primary reason for reduction in grasslands is the increase in spread of 
area under P. juliflora. The model runs indicate that the area under P.juliflora 
will reach 198,000 hectares by year 2030. 



[47] Mihir Mathur and Kabir Sharma 

 

Figure 8: Base Case Land Use Change. All figures in hectares: 1992–2030 

 

Figure 9: Base Case Net Livestock Income: 1992–2030 

 
The period 2004–2008 shows a dip in area under P. juliflora and an increase 
in area under grassland. This is due to the lifting of the ban on charcoal 
making which caused an escalation in removal of P. juliflora from the roots. 
Because of this, the grasses recovered, increasing the area under grassland. 
After the ban was again imposed, it led to growth in area under P. juliflora 
while the grasslands continued to shrink. 

Our base case simulation runs indicate that the net livestock income is 
projected to fall in future years. The decline in net livestock income is 
mainly due to falling livestock population and increase in livestock input 
costs, mainly feed and fodder (due to an increased fodder deficit). These 
input costs spike due to fodder deficit which increases in the later years due 



Ecology, Economy and Society–the INSEE Journal [48] 

 

Figure 10: Fodder Deficit – Base Case 

 

to reducing area under grassland. The input costs are projected to go up 
mainly because of increase in external inputs of feed to compensate for the 
fodder deficit. 

Grassland biomass depends on the extent of rainfall and grassland 
productivity. The variation in rainfall greatly influences the extent of 
grassland productivity and ultimately how much grass grows in that 
particular year. As can be seen in Figure 10 the fodder deficit is expected to 
spike and rise in future years. This is mainly due to reducing grassland area 
coupled with some low rainfall years which lead to low grass production. 
The future trend indicates increase in fodder deficit. It is worth noting that 
in the future years the fodder deficit is never able to fall back to zero as 
seen in past years. This is a cause of concern because it puts continuous 
pressure on Maldharis to buy feed and fodder from outside Banni thereby 
steadily increasing the input costs for livestock maintenance. 

The base case runs present a sorry picture for the livestock economy of 
Banni. If the current spread of P. juliflora continues then the area under 
grassland could reduce to the point that livestock rearing becomes 
uneconomical for the Maldharis of Banni. This could be detrimental since 
livestock forms majority of the income of Banni. Moreover, the loss of 
these fragile grasslands would have numerous other impacts-for 
biodiversity, for biodiversity-based ecotourism and possibly for bird 
migration as well. Also, since it is a low rainfall region, finding alternative 
land-based livelihoods which can compensate for livestock income loss 
could be very difficult. 



[49] Mihir Mathur and Kabir Sharma 

 

Figure 11: Total Livestock under P. juliflora removal policy 

 

Figure 12: Land Use under P. juliflora removal policy 

 

3.2. Policy testing: P. juliflora removal 

Against this backdrop, we have modelled the impacts of a potential P. 
juliflora removal policy (PRP) either decided by the community or the 
government. The P. juliflora area removal rate is kept at 20 per cent per 
annum and the policy becomes active from year 2016 and takes full effect 
after a delay of 3 years. In this scenario the livestock population is estimated 
to increase between 2016–2030 (Figure 11). The main cause for the rise in 
livestock population is the increased fodder availability due to increase in 
area under grassland (Figure 12) due to removal of P. juliflora, increasing 



Ecology, Economy and Society–the INSEE Journal [50] 

 

Figure 13: Fodder Deficit under P. juliflora removal policy 

 

retention of livestock over sales and migration. Also removal of P. juliflora 
reduces the death multiplier on Kankrej, allowing the Kankrej cattle to 
grow more. 

It is projected that the area under grassland would go up to 184,000 
hectares by 2030 while the area under P. juliflora would reduce to 41,000 
hectares and continue to fall. This would increase the grass availability 
leading to an increase in Banni’s livestock carrying capacity. A key 
assumption is that grassland area currently occupied by P. juliflora still has 
grass seeds and that in event of complete removal of P. juliflora the grasses 
would start growing almost immediately. This was observed to happen in 
2004–2008, and nearly all the Maldharis we interviewed believe that this is 
indeed the case. 

Under the PRP scenario the net livestock income is projected to increase 
after a steep dip in year 2020. This increase is mainly attributable to increase 
in area under grassland and subsequent rise in availability of fodder (seen in 
Figure 13). As can be seen from Figure 13, the fodder deficit variable 
returns to zero for multiple years, thus the fodder availability from Banni 
itself is able to sustain the livestock, making livestock rearing more 
profitable by reducing costs. This leads to rise in livestock population due 
to increased livestock carrying capacity while the input costs remain low due 
to abundant fodder availability. Increased livestock leads to increase in milk 
output, dung income, and income from livestock sale, all leading to 
increases in net livestock income (Figure 14). 



[51] Mihir Mathur and Kabir Sharma 

 

Figure 14: Net Livestock Income under P. juliflora removal policy 

 

Figure 15: Total Livestock Population Projections 

 

3.3. Comparing the Scenarios 

Here, the two scenarios are superimposed on each other to give a 
comparative picture, as shown in Figures 15 and 16. 

 

 

 



Ecology, Economy and Society–the INSEE Journal [52] 

 

Figure 16: Net Livestock Income Projections (INR) 

 

As can be seen in figures 15 and 16, P. juliflora removal has a positive impact 
on the livestock population of Banni, mainly due to grassland area 
regeneration. The net livestock income levels also increase. 

In the following section we discount the future earnings to calculate the 
cumulative present value under base case and PRP scenario from year 
2015–2030 using a discount rate of 10 per cent. The difference between the 
two can be assumed to be the partial costs of grassland degradation induced 
by P. juliflora spread in the Banni1. 

3.4. Economic Valuation of Income Flows from 2015 to 2030 

As per the base run model results the total net annual income of Banni in 
year 2015 is around INR 250 crores. The net livestock income, under the 
base case, is projected to continuously decline. The sum of present value of 
livestock and total net income from 2015–2030 comes to INR 1,295 crores 
and INR 1,500 crores respectively (ref. Table 1). If PRP is in place then the 
PV (Present Value) increases to INR 2,230 crores and INR 2,432 crores. 
This indicates that P. juliflora removal has a positive multiplier of 1.7 on the 
economy of Banni. Here it is assumed that the P. juliflora removed is not 
used for charcoal making which if it was then the net impact could have 
been even greater. 

 
1 We assume that these are the partial costs, because we do not include other costs such as of 
loss of biodiversity, loss in tourism incomes and other ecosystem services provided by the 
grasslands. 



[53] Mihir Mathur and Kabir Sharma 

 

One more policy run is done to test the impact of a five year delay in the 
decision to remove P. juliflora and the impact this would have on the PV. 

Table 1: Economic Analysis of Base Case and Policy Runs 

Sr. 
No. 

Present Values (10 per cent 
Discount Rate) 

Net Livestock 
Income 

Net Total 
Income 

1. Base Case 
INR 
12,950,355,676  

INR 
15,019,191,894  

2. 
P. juliflora removal policy (PRP) @ 
20 per cent p.a. 

INR 
22,304,942,323  

INR 
24,326,493,926  

3. Policy Multiplier (PRP÷base Case) 1.7 1.6 

4. 
Difference i.e. costs of grassland 
degradation (No. 2 minus No. 1) 

INR 9,354,586,647  
INR 
9,307,302,032  

5. 
Per ha costs of grassland 
degradation (No. 4÷2,50,000 ha) 

INR 37,418  INR 37,229  
 

Table 2: Economic costs of Policy Delay 

Sr. 
No. 

Present Values (10 per 
cent Discount Rate) 

Net Livestock 
Income 

Net Total Income 

1. PRP with 5 year delay  INR 16,914,549,394  INR 18,984,784,140  

2.  Loss due to delay  -32 per cent -28 per cent 

 

The costs of delaying the implementation of P. juliflora removal policy are 
substantial. The PV for net livestock income comes down by 32 per cent 
while the total net income comes down by 28 per cent due to the delay in 
policy implementation (ref. Table 2). This indicates that PRP is a time 
sensitive policy decision and any delays would result in economic losses for 
Banni. 

3.5. Sensitivity Analysis 

Three rainfall scenarios are modelled to test the sensitivity of rainfall on the 
overall results of the study. In the first scenario we assume that the rainfall 
for 2016 to 2030 would remain same as 2001 to 2015. In the second 
scenario we introduce stochasticity (using the RANDBETWEEN function 
in MS Excel) in the future rainfall keeping the range of the minimum and 
maximum values same as observed between 1992 and 2015, the historical 
duration of the simulation model. In the third scenario we use the min and 
max values from the range of rainfall data from year 1901 to year 2015. The 
data for year 1901 to 2002 is taken from India Water Portal (India Water 
Portal, n.d.). The minimum value is lower while the max value is higher in 



Ecology, Economy and Society–the INSEE Journal [54] 

 

Figure 17: Grass Productivity Sensitivity Runs 

 

this case as compared to the other scenarios, while the mean value is higher. 
Thus, this scenario could be considered as an overall higher mean rainfall 
scenario. The results are discussed below (ref. Table 3). The values for 
annual rainfall for the three scenarios are given in the Annexure. 

The sensitivity analysis reveals that rainfall variability has an impact on the 
livestock net income and overall income of Banni, and the income levels go 
up in the third scenario which is a high rainfall scenario. But the PRP policy 
multipliers remain almost the same, thereby not impacting the policy 
analysis significantly. Same is true in case of testing the impact on PRP 
delay. Thus, rainfall variability does not change our policy 
recommendations even though the model shows sensitivity towards the 
rainfall parameter. 

Further, Figure 17 shows the sensitivity of net livestock income towards the 
grass productivity parameter. Three sensitivity tests are performed, 1) BAU, 
2) high grass productivity and 3) low grass productivity. As can be seen the 
net livestock income shows sensitivity towards the grass productivity 
parameter. At a higher grass productivity, the net livestock income increases 
(red line) while with a lower grass productivity, the income is lower (pink 
line). However, the shape of change remains the same i.e. growth, peak, and 
then decline. This indicates that the model is sensitive towards grass 
productivity, but variation in this parameter does not alter the shape of 
future projections.  



 

Table 3: Present Value Analysis of Rainfall Scenarios 

Sr. 
No 

Particulars 

Scenario 1 Scenario 2 Scenario 3 

Net Livestock 
income 

Net Total 
Income 

Net Livestock 
income 

Net Total 
Income 

Net Livestock 
income 

Net Total 
Income 

1 Base Case 
INR 

12,950,355,676  
 INR 

15,019,191,894  
 INR 

13,572,232,336  
 INR 

15,555,073,054  
 INR 

14,504,560,716  
 INR 

16,487,401,434  

2 
P. juliflora 
removal policy 

 INR 
22,304,942,323  

 INR 
24,326,493,926  

 INR 
24,412,126,383  

 INR 
26,394,967,102  

 INR 
26,675,015,207  

 INR 
28,657,855,926  

3 
Policy 
Multiplier (2÷1) 

1.72 1.62 1.80 1.70 1.84 1.74 

4 
5 year delay in 
PRP Policy 

 INR 
16,914,549,394  

 INR 
18,984,784,140  

 INR 
17,818,453,327  

 INR 
19,801,294,045  

 INR 
18,957,456,833  

 INR 
20,940,297,551  

5 Difference -31.9 per cent -28.1 per cent -37.0 per cent -33.3 per cent -40.7 per cent -36.9 per cent 

 



 

4. DISCUSSION 

The general perceptions of the people of Banni, on the reason for grassland 
degradation point to the growth of area under P. juliflora. It is also widely 
believed that if the P. juliflora is completely removed then the grasses would 
come back. A majority of Maldharis have indicated their preference to 
remain as livestock breeders and pastoralists because they consider it to be 
their traditional, profitable, and sustainable occupation. Our model results 
are consistent with their perceptions and claims. The economic valuation 
indicates that P. juliflora removal is a favourable policy option for sustaining 
their livestock economy and halting grassland degradation. The results 
indicate that livestock profitability goes up in event of P. juliflora removal 
and that in order to sustain livestock as the main occupation of Maldharis 
the land area under P. juliflora needs to be cleared. However, our results 
cannot verify their claims because the model presents a simplified 
representation of Banni. 

The model provides a glimpse into the future possibilities that exist for 
Maldharis and the landscape of Banni based on the use of plausible 
assumptions and parameters. Rainfall is a key variable that determines grass 
productivity, so variation in rainfall could also change the income dynamics, 
as is seen through the sensitivity analysis. This is particularly important for 
Banni since the livestock sensitivity to grass availability is very high and P. 
juliflora density greatly influences the grass availability. 

4.1. Cost of P. juliflora Removal 

There has only been small scale intended P. juliflora removal for grassland 
restoration carried out in Banni, the extent of this being smaller than even a 
village boundary (there are more than 50 villages in Banni). Thus, 
extrapolating the costs of P. juliflora removal based on small scale removals 
done in past to whole of Banni seems inappropriate. Moreover all of those 
efforts were community driven and supported by local NGOs. Thus there 
exists no data on the cost of a formally organized large scale removal 
programme, and thus it becomes hard to account for the effects of 
economies of scale from the data of the small scale removal. However for 
the purpose of this study, an estimate of P. juliflora removal is calculated 
below, taking the example of a small scale removal case. An area of around 
300 sq.m was cleared of P. juliflora by a local NGO with community 
participation in East Banni. The cost of P. juliflora removal was about INR 2 
lac. If we use this example to estimate the cost of P. juliflora removal per 
hectare, a value of INR 6,000 is arrived at. Taking this as a constant value, 
the model simulates the cost of PRP over the period from year 2016 to 
2030 as and discounting (at 10 per cent) to get the present value, a value of 
INR 57 crores is arrived at. Comparing with the increase of INR 935 crores 



[57] Mihir Mathur and Kabir Sharma 

 

in net total income due to PRP, it is evident that the cost of PRP can be 
recovered and that PRP is still a favourable policy option. Moreover, with 
the ready market available for excavated charcoal and wood, the costs could 
further come down. On another point, in the absence of forest rights being 
granted(current scenario), the possibility of community based 
implementation of the decision of large scale P. juliflora removal is 
unrealistic. Thus, unless a government authority decides to carry out a large 
scale removal it is very unlikely that this would happen anytime in the future 
until forest rights are given. 

4.2. Limitations and further scope of research 

This study needs to be strengthened with more data and information about 
the interlinkages between land, biomass, livestock, and economy of Banni. 
There are information gaps with respect to the grass productivity, fodder 
availability in different seasons, extent of seasonal livestock migration due 
to fodder deficit, the role of salinity, charcoal production, future price 
estimates etc. In order to strengthen the results of such a modelling 
exercise, these gaps need to be addressed through empirical field research 
which can then serve as inputs to a further disaggregated system dynamics 
model. There is also the unresolved issue of entitlement of land ownership, 
between the local community, the forest department, and the Revenue 
Department. This makes studying the political ecology of Banni pertinent, 
since these factors would also have a bearing on the decision-making 
processes. 

The current ecological situation of Banni and presence of uncertainty over 
land rights calls for development of decision support tools which can be 
used for performing multi-stakeholder exercises to enable consensual 
decision-making. Thus, this study serves as a motivation for further 
research into the dynamics of the Banni grassland and development policy 
planning tools for consensus development on sustainable management of 
Banni grasslands. 

 

REFERENCES 

Allington, G. R. H., W. Li, and D. G. Brown. 2017. “Urbanization and 
environmental policy effects on the future availability of grazing resources on the 
Mongolian Plateau: Modeling socio-environmental system dynamics.” Environmental 
Science & Policy 68: 35–46. https://doi.org/10.1016/j.envsci.2016.11.005 

Bharwada, C, Mahajan, V. 2012. “Let it be Banni: Understanding and Sustaining 
Pastoral Livelihoods of Banni.” Centre for Economic and Social Studies (CESS) 

https://doi.org/10.1016/j.envsci.2016.11.005


Ecology, Economy and Society–the INSEE Journal [58] 

 

Monograph: RULNR Monograph-13. Begumpet, Hyderabad, India. 
http://www.cess.ac.in/cesshome/mono/CESSMonograph-26(RULNR-13).pdf  

Casti, J. L. 1997. Would-be worlds: How simulation is changing the frontiers of science. New 
York: John Wiley and Sons. 

Dayal, V. 2007. “Social diversity and ecological complexity: How an invasive tree 
could affect diverse agents in the land of the tiger.” Environment and Development 
Economics 12 (4): 553–71. https://doi.org/10.1017/S1355770X07003695 

Directorate General of Census Operations. 2011. Census India. Retrieved 
September 2016, from Census India: 
http://www.censusindia.gov.in/2011census/dchb/2401_PART_B_DCHB_KAC
HCHH.pdf  

FAO. 2005. Plant Production and Protection. Series No. 34. Available at 
http://www.fao.org/docrep/008/y8344e/y8344e00.htm#Contents  

Forrester, J. W. 1958. “Industrial Dynamics -a major breakthrough for decision 
makers.” Harvard Business Review (Jul-Aug). 

Gallati, J., and U. Wiesmann. 2011. “System dynamics in transdisciplinary research 
for sustainable development.” In Research for Sustainable Development: Foundations, 
Experiences, and Perspectives. Perspectives of the Swiss National Centre of Competence in 
Research (NCCR) North-South, University of Bern. vol. 6. ed. U. Wiesmann and H. 
Hurni, 345–360. Bern, Switzerland: Geographica Bernensia. 

Gavali, D., J. Rathod, and T. Patel. 2015. Vegetation dynamics in Banni grasslands under 
the influence of changing climate. Vadodara: Gujarat Ecology Society. 

Geevan, C.P., Dixit, A. M., Silori, C.S. 2003. Ecological Economic Analysis of Grassland 
Systems. EERC Working Paper Series: CPR-5. Bhuj: Gujarat Institute of Desert 
Ecology. http://www.cesc-india.org/doc-archiv/ee-cpg-report-full.pdf 

IMD, n.d. Customized Rainfall Information System. Available at: 
http://hydro.imd.gov.in/hydrometweb/(S(pts0u5ynlm2klf45tyjphx2y))/DistrictRai
fall.aspx), India Water Portal. Met Data. Available at: 
http://www.indiawaterportal.org/met_data/ 

IMD. 2016. Customized Rainfall Information System (CRIS). Retrieved September 
2016, from India Meteorological Department: 
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictR
aifall.aspx 

Koladiya, M.H., Gajera, N. B., Mahato, A.K.R., Kumar, V.V., Asari, R.V. 2016. 
Birds of Banni Grassland. Gujarat Institute of Desert Ecology, Bhuj. The Ravi 
Sankaran Foundation. http://www.ravisankaran.org/wp-
content/uploads/2016/06/Birds-of-Banni-Grassland.pdf 

http://www.cess.ac.in/cesshome/mono/CESSMonograph-26(RULNR-13).pdf
https://doi.org/10.1017/S1355770X07003695
http://www.censusindia.gov.in/2011census/dchb/2401_PART_B_DCHB_KACHCHH.pdf
http://www.censusindia.gov.in/2011census/dchb/2401_PART_B_DCHB_KACHCHH.pdf
http://www.fao.org/docrep/008/y8344e/y8344e00.htm#Contents
http://www.cesc-india.org/doc-archiv/ee-cpg-report-full.pdf
http://hydro.imd.gov.in/hydrometweb/(S(pts0u5ynlm2klf45tyjphx2y))/DistrictRaifall.aspx
http://hydro.imd.gov.in/hydrometweb/(S(pts0u5ynlm2klf45tyjphx2y))/DistrictRaifall.aspx
http://www.indiawaterportal.org/met_data/
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx
http://www.ravisankaran.org/wp-content/uploads/2016/06/Birds-of-Banni-Grassland.pdf
http://www.ravisankaran.org/wp-content/uploads/2016/06/Birds-of-Banni-Grassland.pdf


[59] Mihir Mathur and Kabir Sharma 

 

Kumar, V. V., A. K. R. Mahato, and R. K. Patel. 2015. “Ecology and Management 
of Banni Grasslands of Kachchh, Gujarat.” Ecology and Management of Grassland 
Habitats in India, ENVIS Bulletin. 17: 42-53. 

Mehta, Abhinav; Munmun Sinha and Rohit Chaudhary. 2014. “Evaluation of land 
cover changes in Banni grassland using GIS and RS Technology-A Case Study.” 
Bulletin of Environmental and Scientific Research 3 (4): 18-27.  

MIT. 1997. System Dynamics. Retrieved September 2016, from MIT: 
http://web.mit.edu/sysdyn/sd-intro/ 

Perrings, C. 2010. “Biodiversity, Ecosystem Services, and Climate Change: The 
Economic Problem.” Environment Department Papers, Paper number 120, Environmental 
Economics Series. Washington DC: The World Bank Environment Department. 

Perrings, C., Walker, B. 1999. “Optimal Biodiversity Conservation in Rangelands.” 
Available at http://siteresources.worldbank.org/WBI/Resources/213798-
1104176940274/rangelands.pdf 

Prado, P. D. A., Vázquez, B.E.N., Navarro, R.R.A., and Navarro, R.L.A. 2014. 
Proceedings of the 32nd International Conference of the System Dynamics Society, Delft, 
Netherlands, July: 29–30. 

Rasmussen, L. V., K. Rasmussen, A. Reengberg, and S. Proud. 2012. “A system 
dynamics approach to land use changes in agro-pastoral systems on the desert 
margins of Sahel.” Agricultural Systems 107:56–64. 
https://doi.org/10.1016/j.agsy.2011.12.002 

Rathod, Jagruti Y., and Deepa J. Gavali. 2017. “Decadal study of Avifaunal 
Diversity of Banni Grass land, Katchchh, Gujarat, India.” Biolife. 5 (1): 53–59. 
https://doi.org/10.17812/blj.2017.5110 . 

Reid, R. S., S. Serneels, M. Nyabenge, and J. Hanson. 2005. “The Changing Face 
Of Pastoral Systems In Grass Dominated Ecosystems Of Eastern Africa. ” In 
Grasslands Of The World edited by J. M. Suttie, S. G. Reynolds, and C. Batello. Food 
and Agriculture Organization. 

Sendzimir, J., C. P. Reij, and P. Magnuszewski. 2011. “Rebuilding resilience in the 
Sahel: Regreening in the Maradi and Zinder regions of Niger.” Ecology and Society 16 
(3): art1. https://doi.org/10.5751/ES-04198-160301 

Shah, J. P., A. M. Patel, and P. N. Joshi. 2010. “Participatory natural resources 
mapping: A case study of Bhitara Panchayat in Banni grassland, Kachchh 
(Gujarat).” Asian Journal of Environmental Science 5 (2): 192–98. 

Stephens, E. C., C. F. Nicholson, D. R. Brown, D. Parsons, C. B. Barrett, J. 
Lehmann, D. Mbugua, S. Ngoze, A. N. Pell, and S. J. Riha. 2012. “Modeling the 
impact of natural resource-based poverty traps on food security in Kenya: The 
Crops, Livestock and Soils in Smallholder Economic Systems (CLASSES) model.” 
Food Security 4 (3): 423–39. https://doi.org/10.1007/s12571-012-0176-1 

http://web.mit.edu/sysdyn/sd-intro/
http://siteresources.worldbank.org/WBI/Resources/213798-1104176940274/rangelands.pdf
http://siteresources.worldbank.org/WBI/Resources/213798-1104176940274/rangelands.pdf
https://doi.org/10.1016/j.agsy.2011.12.002
https://doi.org/10.17812/blj.2017.5110
https://doi.org/10.5751/ES-04198-160301
https://doi.org/10.1007/s12571-012-0176-1


Ecology, Economy and Society–the INSEE Journal [60] 

 

Vaibhav, V., Inamdar, A. B., Bajaj, D.N.2012. “Above Ground Biomass and 
Carbon Stock Estimation From P.juliflora in Banni Grassland Using Satellite and 
Ancillary Data.” 33rd Asian Conference on Remote Sensing, Thailand. http://a-a-r-
s.org/acrs/administrator/components/com_jresearch/files/publications/E3-1.pdf  

Vanak, A. T., A. J. Hiremath, S. Krishnan, T. Ganesh, and N. D. Rai. 2017. “Filling 
in the (forest) blanks: the past, present and future of India’s savanna grasslands.” In 
Transcending boundaries. Reflecting on twenty years of action and research at ATREE, edited 
by A. J. Hiremath, N. D. Rai, and A. Siddhartha, 88–93. Bangalore: ATREE.  

Xu, Guangcai, Muyi Kan Kan, Marc Metzger, and Yuan Jiang. 2014. “Vulnerability 
of the Human-Environment System in Arid Regions: The Case of Xilingol 
Grassland in Northern China.” Polish Journal of Environmental Studies 23 (5): 1773–17. 

 

ANNEXURE 

Rainfall Scenarios 

Year Scenario 1 Scenario 2 Scenario 3 

1992 507 507 507 

1993 106 106 106 

1994 729 729 729 

1995 326 326 326 

1996 174 174 174 

1997 259 259 259 

1998 464 464 464 

1999 450 450 450 

2000 195 195 195 

2001 540 540 540 

2002 110 110 110 

2003 700 700 700 

2004 147 147 147 

2005 139 139 139 

2006 485 485 485 

2007 641 641 641 

2008 177 177 177 

2009 370 370 370 

2010 655 655 655 

2011 650 650 650 

2012 350 350 350 

2013 652 652 652 

2014 291 291 291 

2015 450 450 450 

2016 540 397 592 

2017 110 264 433 

http://a-a-r-s.org/acrs/administrator/components/com_jresearch/files/publications/E3-1.pdf
http://a-a-r-s.org/acrs/administrator/components/com_jresearch/files/publications/E3-1.pdf


[61] Mihir Mathur and Kabir Sharma 

 

2018 700 172 250 

2019 147 326 634 

2020 139 718 729 

2021 485 512 483 

2022 641 592 486 

2023 177 330 473 

2024 370 233 307 

2025 655 252 204 

2026 650 566 644 

2027 350 377 720 

2028 652 287 319 

2029 291 472 461 

2030 450 282 220 

Mean Value 408.31 393.51 423.64 

Scenario 1 = Future rainfall (2016 to 2030) same as past (2001 to 2015) 
Scenario 2 = Future rainfall (2016 to 2030) with stochasticity based on 2001 
to 2015 rainfall data 
Scenario 3 = Future rainfall (2016 to 2030) projected with stochasticity 
based on rainfall data from 1901 to 2015( having a higher mean) 

Parameter Values and Sources 

S 
No. 

Parameter Value Taken 
Sources and Explanations 
where necessary 

1.  
Initial Value (1992) 
of Adult Buffaloes  

12,580 

Taken as 75 per cent of the 
total buffalo population in 
1992 (Bharwada and Mahajan 
2012, 87) 

2.  
Initial Value of 
buffalo Calves  

4,194 

Taken as 25 per cent of the 
total buffalo population in 
1992 (Bharwada and Mahajan 
2012, 87) 

3.  
Fraction of adult 
buffaloes giving 
birth every year 

0.5 
Discussions with local NGO 
Sahjeevan and Maldharis 

4.  
Buffalo calf death 
rate 

20 per cent p.a. 
Data from personal interview 
with experts and pastoralists 

5.  
Buffalo calf 
maturation time 

3 years 

Discussions with local NGO 
Sahjeevan and Maldharis 

6.  
Normal buffalo sale 
rate  

1 per cent p.a. 

7.  Buffalo lifetime 
23 years (3 yrs. as 
calf and 20 as 
adult) 



Ecology, Economy and Society–the INSEE Journal [62] 

 

8.  
Fodder requirement 
per adult buffalo 
per day 

30 kg 

9.  
Fodder requirement 
per buffalo calf per 
day 

7.5 kg 

10.  
Fraction of milk 
producing buffaloes 

50 per cent 

11.  
Initial Value (1992) 
of Adult Kankrej 

4,544 
Taken as 75 per cent of the 
total Kankrej population in 
1992 (Bharwada and Mahajan 
2012, 87) 

12.  
Initial Value (1992) 
of Kankrej Calves 

1,514 

13.  Kankrej birth rate 

50 per cent of 
adult Kankrej 
cattle give birth 
every year 

Discussions with local NGO 
Sahjeevan and Maldharis 

14.  
Kankrej calf death 
rate 

20 per cent p.a. 

15.  
Average Kankrej 
calf sale rate 

60 per cent p.a. 

16.  
Average male 
Kankrej purchase 
rate 

25 per cent p.a. 

17.  
Kankrej calf 
maturation time  

3 years 

18.  Kankrej lifetime 
12 years as adult 
and 3 years as calf 

19.  
Fraction of milk 
producing Kankrej 

50 per cent p.a. 

20.  
Fodder requirement 
per Kankrej adult 
per day 

15 kg 

21.  
Fodder requirement 
per Kankrej calf per 
day 

5 kg 



[63] Mihir Mathur and Kabir Sharma 

 

22.  
Buffalo sale 
multiplier due to 
profitability 

Increases from 0 
to 30 per cent with 
profit per livestock 
falling from INR 0 
to INR -5000. 

Parameterized using sensitivity 
runs: Variables that are 
difficult to elicit through a 
survey or field estimation are 
determined through a series of 
plausibility tests. Parameters 
for such variables are assumed 
and tested through 
simulations. If the simulation 
outcome shows high deviation 
from real life trends then the 
parameters are optimized 
through sensitivity runs until 
the simulations generate 
plausible/verifiable with past 
data behaviour over time 

23.  
Kankrej sale 
multiplier due to 
profitability 

Increases from 0 
to 20 per cent with 
profit per livestock 
falling from INR 0 
to INR -5000. 

Parameterized using sensitivity 
runs 

24.  
Impact of P. juliflora 
on death rate of 
Kankrej 

Increases from 0 
to 20 per cent and 
tapers off as P. 
juliflora density 
doubles 

Parameterized using sensitivity 
runs 

25.  Rainfall 

Rainfall from 
2015–2030 
assumed to be the 
same as from 
1999–2014. 

Rainfall data for 1992–2010 
taken from (Bharwada and 
Mahajan 2012, 143), for year 
2011–12 taken from, (Gavali 
2015, 5) and for 2013–14 
taken from IMD website for 
Kutch district from 
http://hydro.imd.gov.in/hydr
ometweb/(S(lmae0jvse31sb04
5m2gxd5i1))/DistrictRaifall.as
px 

26.  
Initial Value (1992) 
of Area under P. 
juliflora 

41180 ha 

Koladiya et al. (2016, 20) 

27.  
The total 
productive area of 
Banni 

225,000 hectares  

28.  
Normal spread rate 
of P. juliflora 

8.5 per cent Vaibhav et. al (2012) 

http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx
http://hydro.imd.gov.in/hydrometweb/(S(lmae0jvse31sb045m2gxd5i1))/DistrictRaifall.aspx


Ecology, Economy and Society–the INSEE Journal [64] 

 

29.  
Impact of livestock 
on P. juliflora spread 

Increasing from 1 
to 2 when 
livestock 
population 
increases from 
25,000 to 100,000 

Parameterized using sensitivity 
runs 

30.  
Charcoal 
production 

4,800 sacks of 
40 kg each 
produced per day  

Discussions with local NGO 
Sahjeevan and Maldharis 

31.  
Impact of profit per 
livestock on 
charcoal production 

As profit per 
livestock falls 
below 0, this 
function begins to 
increase from 1 
and goes up until 2 
at a loss of INR -
5000 per livestock 

Parameterized using sensitivity 
runs 

32.  
Average milk 
production per 
buffalo per day 

12 litres 

Discussions with local NGO 
Sahjeevan and Maldharis Milk 
production per buffalo ranges 
from 8 litres to 20 litres a day. 
Average taken as 12 litres a 
day.  

33.  
Milk price per litre 
of Banni buffalo 
milk 

Graphical function 
varying from Rs.19 
per litre in 1992 to 
Rs. 40 per litre in 
2015. Kept at 2015 
prices in future. 

Historical milk prices taken at 
2015 constant values. 
2015 milk price taken from 
personal interviews with dairy 
industry. 
2010 milk price taken from 
Bharwada and Mahajan 2012, 
71) 
2000 milk price taken from 
(Geevan et al. 2012, 56) table 
6.9 
1992 milk prices are assumed. 

34.  
Average milk 
production per 
Kankrej per day 

9 litres 

Discussions with local NGO 
Sahjeevan and Maldharis Milk 
production per Kankrej cattle 
ranges from 6 to 14 litres a 
day. Average taken as 9 litres a 
day. 



[65] Mihir Mathur and Kabir Sharma 

 

35.  
Milk price per litre 
of Kankrej cattle 
milk 

Graphical function 
varying from Rs.10 
per litre in 1992 to 
Rs. 18 per litre in 
2015. Kept 
constant at 2015 
prices in future. 

Historical milk prices taken at 
2015 constant values. Current 
prices for 2015 taken from 
personal interview, while 
earlier prices are re-calculated 
to reflect 2015 constant values 

36.  Charcoal Price 
Rs. 5/kg taken 
constant  

Discussions with local NGO, 
Sahjeevan, Personal interviews 
with Maldharis 

37.  Price of Dung 
Rs 1,500 per truck 
load 

Bharwada and Mahajan (2012, 
74) 

38.  
Quantity of Dung 
sold 

One truck load 
every 15 days- one 
truck load from 
100 livestock 

39.  Kankrej sale price Rs 10000 

Average price varies from Rs 
12,000 to Rs 30,000 for a pair 
of bullock. Taken as average 
Rs. 10000 per Kankrej. 
(Bharwada and Mahajan 2012, 
65) 

40.  Buffalo sale price 

Varying from Rs 
38,000 in 1992 to 
Rs75000 in 2015 
(post-breed 
registration). 
Constant at Rs 
75,000 in future. 

Current buffalo price for year 
2015 range from INR 50,000 
to INR 300,000. Mode value 
of sale price taken as INR 
75,000 and then normalized 
for the past years taking into 
consideration the rise in price 
due to buffalo registration in 
year 2011 

41.  
Input cost for milk 
producing buffaloes 

Graphical function 
of fodder deficit. 
Varies from 10000 
at 0 fodder deficit 
to 140,000 at 100 
per cent fodder 
deficit  

At 50 per cent fodder deficit 
the cost of feed for milk 
producing buffalo is estimated 
to be Rs. 70,000/- per annum. 
The numbers are adjusted to 
reflect fall and increase in 
fodder deficit and its 
corresponding impact on feed 
cost due to increase in supply  

42.  
Feed cost for non-
milk producing 
buffaloes 

One-third of No. 
36. 

Discussions with local NGO 
Sahjeevan and Maldharis 

 


