Introduction

Land use land cover (LULC) dynamics analyses help in understanding the impact of anthropogenic pressures on Earth system processes (Lambin, Geist, & Lepers, 2003; Ramachandra, Setturu, & Bharath, 2012). Forest ecosystems constitute a key component of the global carbon cycle that accounts for over two-thirds of net primary production on land through photosynthesis, converting solar energy into biomass. Forests support local livelihoods through the provision of non-timber forest products (NTFPs), fuel, wood, water and so on. Unplanned developmental activities such as illegal logging, mining, the unsustainable exploitation of resources and so on have been posing serious threats to the sustenance of water resources. This is also evident from the escalation of human-animal conflicts, water scarcity and so on. Deforestation, one of the prime movers of global warming, and consequent changes in the climate are prevalent throughout the globe. Irrational land cover changes resulted in the loss of more than one-thirds of global forest cover in the post-industrialization and globalization era (Schneider, Friedl, & Potere, 2009; Sexton et al., 2013). LULC changes are highly interlinked processes such as the intensification of agriculture, tropical deforestation, pasture expansion, urbanization and so on; they have a wide range of impacts on livelihood, net primary production, biogeochemical cycles, ecosystem stability and biodiversity through land surface processes (Akber & Shrestha, 2015; Haase & Nuissl, 2010). Human-induced LULC changes are the major drivers of the landscape dynamics at local levels. LULC changes alter the homogeneous landscape into a heterogeneous mosaic of patches. This leads to the division of habitat into smaller and more isolated patches known as fragmentation (Liu, Feng, Zhao, Zhang, & Su, 2016; Ramachandra, Setturu, & Chandran, 2016). Fragmented forests are likely to suffer from being smaller and more isolated with greater edges which alters the structure of the landscape, affecting ecosystem functional abilities, which is evident from the decline of biodiversity, hydrology and so on (García-Llorente, Martín-López, Nunes, Castro, & Montes, 2012; Kang, Zhu, Zhu, Sun, & Ou, 2014; Vinay, Bharath, Bharath, & Ramachandra, 2013). This induces change in the local climate (Bharath, Rajan, & Ramachandra, 2013; Ramachandra, Bharath, & Sreejith, 2015) as well as in the global climate (Chase, Pielke Sr, Kittel, Nemani, & Running, 2000; Foley et al., 2005). LULC changes have thus gained special attention for land management and planning due to their potentially negative consequences, creating trade-offs between ecosystem services (Polasky, Nelson, Pennington, & Johnson, 2011; Schägner, Brander, Maes, & Hartje, 2013).

Inventorying, mapping and the monitoring of biological resources through LULC analysis aid in the conservation of ecologically rich regions through an understanding of spatial patterns in the environment. The advancement of technologies in vegetation mapping and geospatial models has provided for a precise evaluation of spatio-temporal patterns of forest dynamics and environmental consequences. Remote sensing (RS) data with advancements in spectral, spatial, radiometric and temporal resolutions and geographical information system (GIS) techniques have been useful in monitoring land use changes of even inaccessible areas within a short span. Modelling and the visualization of landscape dynamics help in understanding the status of natural resources, considering the environmental and socio-economic processes, which facilitate in prudent resource planning with scientific insights to meet or mitigate potential threats. Rapid advances in the availability of multi-resolution spatial data and geospatial models have made it increasingly possible to design and simulate spatial patterned LULC changes (Bharath, Vinay, & Ramachandra, 2014; Ramachandra, Bharath, & Bharath, 2014; Verburg, Schot, Dijst, & Veldkamp, 2004). The proven modelling techniques such as cellular automata (CA), Markov chains, multivariate statistics, optimization, system dynamics, multi-criteria evaluation (MCE) and the conversion of land use and its effects model (CLUE) aid in visualizing complex patterns and predictions. The Markov model is based on a stochastic theory of random process systems and the optimal control theory (Pielke Sr et al., 2011; Rands et al., 2010). The traditional Markov model alone is difficult to predict the spatial pattern of land use changes. The CA model has a strong space conception, which is strongly capable of space-time dynamic evolution with complex space systems. The integrated CA-Markov is a robust and effective modelling approach to analyse the spatio-temporal patterns of LULC changes (Bharath, Rajan, & Ramachandra, 2016; Jiang, Zhang, & Kong, 2009; Kamusoko, Aniya, Adi, & Manjoro, 2009). The spatial simulation modelling technique such as CA-Markov is an efficient tool for quantitatively exploring forest cover change (Ramachandra, Bharath, & Gupta, 2018). The spatial modelling technique such as CA-Markov focuses mainly on local interactions of cells with distinct spatio-temporal features suitable for dynamic simulation and display (Wu, Ge, & Dai, 2017). CA-Markov modelling comprises observations and a structural concept of various forces and functions at various scales in a specific forested landscape (Vázquez-Quintero et al., 2016). CA-Markov incorporates rules based on the consideration of biophysical and socio-economic data to define initial conditions, parameterize the CA-Markov model and calculate transition probabilities and the neighbourhood rules with transition potential maps. The transition rules of CA models help in accounting the temporal and spatial complexities of forest systems. Bannerghatta National Park (BNP) is located in close proximity to Bangalore (20 km to the BNP core area and

• 1. km from the Bruhat Bengaluru Mahanagara Palike (BBMP) boundary), capital city of Karnataka. The peri- urban region of BBMP growth centres and growth poles play a key role in initiating the process of industrialization and urbanization. The growing uncontrolled economic activities in and around the BBMP region threaten local well-being, agricultural regions and the forests of BNP. The objectives of current study are:
     1. assessment of the status of forests in and around ecologically sensitive regions, BNP with a 5-km buffer from 1973 to 2027;
     2. understanding the transitions in LULC with the causal factors from 1973 to 2015;
     3. assessment of the spatial patterns of forest changes from 1973 to 2015; and
     4. the suggestion of prudent management options (the restoration of degraded forest patches and sustainable management) to mitigate forest loss.