Hi, I am Rounak Choudhary and welcome to this bolgpost where i tried to simplify and summerise my doctoral research so that every person can understand it with an ease (Currently it is in developmental stage). My thesis investigates the relationships between landscape heterogeneity, vertebrate biodiversity, functional diversity and wildlife habitat capacity of agroecosystems in Rajasthan, India. Grounded in landscape ecology, my study evaluates how human land use, spatial configuration and environmental filtering (such as aridity) interact to shape ecological communities in human-dominated landscapes.

You can also read about my work on Urban Ecology, Raptor Ecology and SDM of Endemic and Threaten Species.

The PDF of thesis will be available here.

The reproducible codes and workflows will be available here.

To Understand Nature, You Must See the Whole Tapestry

The foundation of this perspective lies in the discipline of landscape ecology, which demands a wider lens than traditional biology. Rather than focusing on a single plant or an isolated population, we must shift toward "holism." This approach recognizes that an ecosystem's health is born from the constant interaction between ecological, social and cultural dimensions.

The Secret Geometry of Survival: Why the Way We Arrange Farms Matters More Than You Think

In our modern imagination, we tend to categorize the natural world as a series of isolated retreats, neatly fenced national parks or remote wilderness areas separated from the "unnatural" sprawl of human activity. We view the farm as a factory for food and the forest as a biological fortress, with little meaningful connection between the two.However, ecological studies suggests this isolation is a dangerous illusion. Species do not exist in vacuums; they inhabit a "landscape structure" that governs their survival across vast spatial scales. To understand how life persists in a human-dominated world, we must look at the matrix, the often-overlooked space between those protected islands. If nature is a complex puzzle, we are realizing that the geometry of the pieces matters just as much as what is printed on them.

The Carl Troll Rule: Nature as a Whole

This holistic view was coined by German geographer Carl Troll, who formally coined the term "landscape ecology" in 1939. Drawing on a European tradition that views human activity and natural processes as an inseparable unit, he famously stated: "To understand nature, one must see the landscape as a whole". This perspective prevents the critical mistake of managing only one fragment of an ecosystem. It recognizes that landscapes are "multifunctional" systems—dynamic, scale-dependent, and constantly reshaped by human land-use.

The "What" vs. The "Where": Composition vs. Configuration

When measuring the health of a landscape, ecologists look at two distinct features: composition and configuration. The distinction is the difference between a functional ecosystem and a biological collapse.

Geometry vs. Ingredients: Why Two Identical Landscapes Can Fail Differently

Landscape Composition describes the "what"—the identity and variety of elements present, such as the percentage of cropland, natural vegetation or water bodies. It is essentially the inventory of the biological pantry. Landscape Configuration , however, is the "where." It refers to the spatial arrangement, geometry and connectivity of those ingredients. This includes patch size, edge density and the degree of fragmentation.

This leads to a counter-intuitive insight: two landscapes can have the exact same amount of habitat (identical composition), yet support vastly different levels of life. In a poorly configured landscape, habitat is shattered into tiny, isolated islands, cutting off gene flow and disrupting species interactions. Without connectivity, even a resource-rich patch becomes a dead end. When corridors allow for movement, that same amount of land becomes a functional highway for survival.

The Farm is a Mosaic, Not a Factory

We often mistake agricultural lands for biological deserts, but landscape ecology reframes "agroecosystems" as vital ecological mosaics.

The primary objectives of the study were:

But why Rajasthan?

The Master Sculptor: How Aridity Dictates Life in Rajasthan

In Rajasthan, ecological survival is measured on a concrete scale of Aridity, which ranges from the hyper-arid Thar Desert to the sub-humid plateaus. The environment is defined by extreme temperature fluctuations (frequently exceeding 45°C) and high evapotranspiration. In this high-stress system, the "temporal pulse" of the monsoon is the heartbeat of the land. During the monsoon, productivity peaks, but the real test of the landscape's geometry occurs during the post-monsoon and dry seasons. In these periods, vegetation cover shrivels and water disappears. "Localized zones of enhanced resource"—such as irrigated patches—function as vital temporal anchors. These patches keep the system from collapsing during the lean months, providing refuge for vertebrates when the wider matrix becomes inhospitable.

Mean annual temperature ranges from under 8°C hill stations to over 45°C in the hyper-arid regions of the Thar Desert.

Mean annual rainfall ranges from under 200 mm in the hyper-arid west to over 600 mm in the sub-humid southeast, with monsoon precipitation creating strong seasonal resource pulses.

The Aravalli Hill range runs center-state, creating significant topographic and soil variation.

Thar Desert: Hyper-arid, sandy terrain with xerophytic vegetation and sparse scrub. Eastern Desert Margin: A transitional ecotone with mixed sandy/loamy soils and scattered agriculture. Aravalli Hills: Ancient rocky range with dry deciduous/thorn forests providing wildlife refugia. Semi-Arid Plains: Fertile agricultural plains with extensive cropping and scattered tree cover. Sub-Humid Plateau: Wetter region with higher vegetation productivity and plateau terrain.

The Methodology

Studying biodiversity across an entire state is no small task. We needed a sampling framework large enough to capture landscape-scale ecological processes while remaining practical for field surveys.

Twenty-five study landscapes were established across Rajasthan, each measuring 25 × 25 kilometres. Every grid was treated as an independent landscape unit for ecological analysis. These landscapes were distributed across the five ecological regions: Thar Desert – 7 sites; Eastern Desert Margin – 4 sites; Aravalli Hills – 4 sites; Semi-Arid Plains – 5 sites; Sub-Humid Plateau – 5 sites. To further capture climatic variation, the sites were categorized into arid and semi-arid zones. Thirteen sites represented arid conditions, while twelve represented semi-arid environments. Care was also taken to distribute sampling locations evenly across the state to minimize spatial bias and maximize geographic representation.

Why Vertebrates?

Biodiversity encompasses everything from microorganisms to trees, insects, fungi and mammals. So why focus specifically on vertebrates?

The answer lies in their ecological importance and sensitivity to environmental change. Vertebrates occupy multiple trophic levels and perform diverse ecological roles, including herbivory, predation, scavenging, seed dispersal and nutrient cycling. Their responses often reflect broader changes occurring within ecosystems.

The study concentrated on four vertebrate groups: Birds; Mammals; Reptiles; Amphibians.

These groups are particularly useful because they: Respond strongly to habitat alteration; Are relatively well documented taxonomically; Possess extensive ecological trait information; Serve as indicators of habitat quality and environmental change.

Between January 2022 and January 2026, vertebrate surveys were conducted across all study landscapes using standardized strip transect sampling. A strip transect involves moving along a predefined route and recording all animals observed within a specified distance from the survey line. In this study, surveys used an effective strip width of 50 metres.

This method offered several advantages: Standardized sampling effort; Efficient coverage of large areas; Reliable abundance estimates; Comparability across landscapes

However, wildlife does not always cooperate with ecological surveys. Different species vary greatly in visibility, behaviour, and activity patterns. To maximize detection, We employed multiple survey techniques.

1) Watching Wildlife: We conducted visual encounter surveys using binoculars and direct observation, identifying animals based on morphology, coloration, body size, and behaviour.

2) Listening to Nature: Birds and some amphibians were detected through their vocalizations using field guides and reference call libraries. This acoustic approach allowed us to record species that remained hidden from view.

3) Following the Clues: Wildlife often leaves evidence even when the animals themselves remain unseen. Tracks, footprints, burrows, feeding signs, scat, and shed skins were used to confirm species presence, particularly for elusive and nocturnal animals.

Measuring Biodiversity: More Than Counting Species

A landscape containing fifty species is not necessarily more diverse than one containing forty. What matters is also how individuals are distributed among those species. To capture these nuances, We calculated a suite of biodiversity metrics. Species richness measured the total number of species present. Dominance quantified whether a few species disproportionately controlled the community. Simpson and Shannon diversity indices incorporated both richness and abundance distributions, providing more sophisticated measures of community structure. Evenness assessed how equally individuals were distributed among species, while the iChao-1 estimator accounted for species likely missed during sampling.

Looking Beyond Species: Functional Diversity

Suppose two landscapes each contain twenty bird species. At first glance, they appear equally diverse. But what if one landscape contains species that feed in trees, forage on the ground, disperse seeds, control pests, and occupy many ecological niches, while the other contains species performing largely similar functions?

This distinction is captured by functional diversity. Functional diversity focuses on what species do rather than simply how many species are present. It measures ecological strategies, roles, and traits that influence ecosystem functioning. To assess functional diversity, We compiled trait information from globally recognized databases: AVONET; EltonTraits; AmphiBIO; ReptTraits.

Understanding the Landscape Itself

Wildlife does not exist independently of its surroundings. To understand biodiversity patterns, We needed to characterize the landscapes in which species lived. For this purpose, annual land-cover maps derived from the Esri Sentinel-2 Land Cover dataset were used. These high-resolution maps provided detailed information about landscape structure from 2017 to 2024.

Seven major land-cover categories were identified: Water bodies; Tree cover; Wetlands; Farmland; Built areas; Bare ground; Rangelands

Each category contributes differently to habitat availability, resource distribution, and wildlife movement.

Quantifying Landscape Heterogeneity

The term "landscape heterogeneity" is often used in ecology, but what does it actually mean? In simple terms, it describes how diverse and spatially complex a landscape is. We quantified heterogeneity using a broad suite of landscape metrics. These metrics measured both composition and configuration. Composition metrics described the diversity and proportional representation of land-cover classes. Examples included patch richness, Shannon diversity, and Simpson diversity. Configuration metrics described how those land-cover classes were arranged in space. These included measures of fragmentation, connectivity, edge complexity, patch shape, and isolation.

Simplifying Complexity

With dozens of landscape metrics available, interpreting ecological patterns can become difficult. To overcome this challenge, we used Principal Component Analysis (PCA), a statistical technique that condenses many correlated variables into a smaller number of meaningful dimensions. This process produced nine synthetic landscape indices representing major aspects of landscape structure: Fragmentation; Patch Size; Core Area; Patch Density; Edge Complexity; Shape Complexity; Isolation; Spatial Mixing; Composition.

Creating an Index for Wildlife Habitat Capacity in Agroscapes

The ultimate goal was not simply to describe biodiversity or landscapes independently. Instead, the study sought to evaluate how capable each landscape was of supporting wildlife. To accomplish this, we developed a Habitat Capacity Index (HCI) that integrated three major components:

Biodiversity
Functional Diversity
Landscape Structure

Using PCA-derived indices, these components were combined into a single ecological measure representing wildlife habitat capacity. Rather than focusing on one ecological dimension, the HCI reflected the combined influence of species diversity, ecological functioning, and habitat suitability. This made it possible to compare landscapes across Rajasthan using a standardized measure of ecological potential.

Connecting Landscapes and Biodiversity

The final phase of the study focused on understanding cause-and-effect relationships. How does fragmentation affect biodiversity? Does habitat mixing promote ecological functioning? Are more heterogeneous landscapes better able to support wildlife? To answer these questions, we used multivariate statistical techniques including PCA, Principal Coordinates Analysis (PCoA), Generalized Linear Models (GLMs), and regression analyses. These analyses linked biodiversity patterns directly to landscape characteristics and environmental gradients, revealing how ecological communities respond to changes in landscape structure.

The Results

Rajasthan's Hidden Wildlife Story: What Landscape Change Reveal About Biodiversity in India's Largest State

From desert reptiles and birds to amphibians and mammals hidden in the human-dominated landscapes, Rajasthan's wildlife is constantly responding to changes in the landscape. But what happens when rangelands become farms, farms become towns and habitats become increasingly fragmented?

A State in Transition

Rajasthan is often associated with sand dunes, camels and the vast Thar Desert. Yet beneath this familiar image lies an extraordinary ecological mosaic stretching from arid deserts to semi-humid plateaus, from rocky mountain ranges to productive agricultural landscapes. Over the last decade, these landscapes have undergone substantial transformation. Agricultural expansion, urban growth, infrastructure development, and changing land-use practices have altered the ecological fabric of the state. Understanding how these changes affect biodiversity is essential for conservation planning in a region where wildlife increasingly survives outside protected areas.

Taking a Note of Rajasthan's Wildlife in Agroecosystems

The first step was simple in concept but ambitious in scale: document vertebrate diversity across Rajasthan's agricultural landscapes.

The surveys recorded 226 vertebrate species, representing four major taxonomic groups: 166 bird species; 32 reptile species; 21 mammal species; 7 amphibian species

Birds emerged as the most diverse group by a considerable margin, accounting for nearly three-quarters of all recorded species. Their ability to occupy a wide range of habitats—from agricultural fields and wetlands to scrublands and forests—makes them particularly responsive to landscape heterogeneity. Reptiles showed moderate species richness but relatively even community structures, indicating that no single reptile species overwhelmingly dominated the landscape. Mammals displayed lower species richness but substantial ecological variation across regions, particularly in arid and semi-arid environments. Amphibians were the least diverse group, reflecting Rajasthan's generally dry climate. Yet despite their low species numbers, amphibians revealed some of the most important ecological patterns in the study.

Not All Landscapes Are Equal

One of the clearest findings was that biodiversity varied dramatically among Rajasthan's ecological regions.

The Sub-Humid Plateau: Rajasthan's Biodiversity Hotspot

The Sub-Humid Plateau consistently emerged as the strongest performer for biodiversity. Higher rainfall, greater vegetation productivity, and increased habitat complexity supported: Highest amphibian abundance; High bird richness; High reptile richness; Greater overall ecological complexity In many ways, this region functioned as Rajasthan's biodiversity engine.

Semi-Arid Plains: Surprisingly Rich

The Semi-Arid Plains also supported high biodiversity. Birds and reptiles were particularly diverse here, likely benefiting from the coexistence of agriculture, scattered natural habitats, and heterogeneous land-use patterns. These findings reinforce a growing realization in conservation science: Productive agricultural landscapes can support substantial biodiversity when habitat heterogeneity is maintained.

Aravalli Hills and Eastern Desert Margin

The Aravalli Hills and Eastern Desert Margin occupied an intermediate position. These landscapes supported moderate levels of biodiversity and often acted as transition zones between highly productive and highly arid ecosystems. The structural complexity provided by hills, rocky terrain, and mixed vegetation likely contributes to their ecological importance.

The Thar Desert: Biodiversity Under Constraint

The Thar Desert recorded lower species richness, particularly for mammals and amphibians. This does not mean the desert lacks ecological value. Rather, environmental filtering is much stronger. Species that persist here are highly specialized and adapted to extreme conditions, making them ecologically unique despite lower overall richness.

Amphibians: Nature's Moisture Sensors

Although amphibians represented the smallest taxonomic group, they proved to be among the most informative. Amphibian communities were strongly influenced by habitat composition and moisture availability. Even small changes in landscape quality produced noticeable responses. Two species stood out: Indian Marbled Toad - This species was strongly associated with less arid, moisture-rich environments and emerged as one of the most reliable ecological indicators in the study. Asian Grass Frog - The Asian Grass Frog showed strong associations with landscapes characterized by structural complexity and irregular patch geometry.

Reptiles: Specialists of Structure and Aridity

Reptiles told a different story. Unlike amphibians, many reptiles were adapted to arid environments and responded strongly to landscape structure. Several species emerged as indicators of environmental conditions. Brilliant Ground Agama - This species was closely associated with highly arid conditions and served as an indicator of desert environments. Garden Skink - The Garden Skink emerged as one of the strongest indicators of low-aridity landscapes. Russell's Viper and Trinket Snake - These reptiles were associated with landscapes exhibiting high compositional heterogeneity, reflecting the importance of habitat mixing. Reptiles demonstrated how structural complexity and environmental gradients jointly shape ecological communities.

Birds: Masters of Landscape Heterogeneity

Birds exhibited the greatest ecological variation among all taxonomic groups. Some species thrived in highly fragmented agricultural systems. Others depended on more complex habitat mosaics. Key indicator species included: Yellow-Eyed Pigeon and Asian Desert Warbler Linked to rangeland-dominated landscapes. Streak-Throated Swallow - Associated with large patch areas and lower patch density. Bird communities effectively captured broad-scale land-use patterns, making them particularly useful for landscape monitoring.

Mammals: Responding to Human Land Use

Mammals emerged as some of the strongest indicators of landscape configuration. Their abundance and distribution were strongly influenced by land-use change and habitat arrangement. Small Indian Civet - One of the most robust mammalian indicators, particularly in landscapes characterized by lower aridity and greater habitat diversity. Rhesus Macaque - Associated with highly fragmented landscapes. Grey Mongoose - Frequently linked to areas exhibiting substantial fragmentation. These patterns suggest that mammals respond strongly to both habitat quality and anthropogenic disturbance.

Functional Diversity: Looking Beyond Species Counts

Biodiversity is not simply about the number of species present. Equally important is what those species do. This concept, known as functional diversity, examines ecological roles and strategies. The study revealed several fascinating patterns:

Mammals Lead Functional Specialization

Mammals displayed the highest levels of functional specialization and dispersion. This indicates the presence of species occupying highly distinct ecological roles.

Amphibians Lead Functional Richness

Despite having the fewest species, amphibians showed the highest functional richness and originality. In ecological terms, each amphibian species contributed disproportionately unique functions. This means losing even a few amphibian species could result in substantial losses of ecological function.

Biodiversity patterns make more sense when viewed alongside changes in land cover. Using high-resolution land-cover data from 2017 to 2024, we reconstructed how Rajasthan's landscapes evolved over time. The results revealed a clear trend:

Rangelands Are Shrinking: In 2017, rangelands occupied approximately 47.6% of the state's area. By 2024, this had declined to about 43.4%. At the same time, farmland expanded and became Rajasthan's dominant land-cover class. This shift reflects the growing conversion of grazing lands into agricultural production systems.

Urban Areas Are Expanding Rapidly: Perhaps the most striking transformation was the growth of built-up land. Urban and developed areas increased by approximately 61% over the study period. While built-up land still occupies a relatively small percentage of Rajasthan's total area, its rapid expansion represents a significant ecological pressure. Urbanization introduces: Habitat loss; Increased disturbance; Road networks; Fragmentation; Reduced connectivity; All of which influence wildlife communities.

Fragmentation Is Increasing; The study found strong evidence that Rajasthan's landscapes are becoming increasingly fragmented. Several indicators pointed toward this trend:

Average patch sizes declined substantially.
Patch density increased.
Landscapes became divided into smaller habitat units.
Edge effects intensified.

In simple terms, habitats are being broken into smaller pieces. This fragmentation emerged as one of the most important drivers of biodiversity patterns across the entire study.

When Bigger Is Not Always Better

One of the most interesting findings challenged a common assumption. Conservation discussions often assume that larger habitat patches are automatically better. However, the study revealed what we called a "Patch Area Paradox." Larger habitat patches were not always associated with higher biodiversity or habitat capacity. In some cases, large patches represented simplified landscapes with lower ecological complexity. Conversely, smaller but structurally diverse habitat mosaics often supported richer ecological communities.

This highlights an important principle of landscape ecology: Habitat quality and heterogeneity can be more important than habitat size alone.

The Universal Threat: Fragmentation

Across nearly every analysis, one pattern remained remarkably consistent. Fragmentation was the strongest negative predictor of ecological performance. Whether examining birds, mammals, amphibians, reptiles, or overall habitat capacity, increasing fragmentation generally reduced ecological quality.

Fragmentation influences biodiversity by:

Limiting movement
Reducing connectivity
Increasing edge effects
Isolating populations
Altering resource availability

The findings suggest that maintaining connected habitat networks may be one of the most effective conservation strategies for Rajasthan's agricultural landscapes.

The Habitat Capacity Index: Measuring a Landscape's Ability to Support Life

One of the most innovative outcomes of the study was the development of a Habitat Capacity Index (HCI). Rather than relying on a single biodiversity metric, the index integrated: Species diversity; Functional diversity; Landscape structure;

The objective was straightforward: Measure how capable a landscape is of sustaining wildlife communities. The analyses revealed that habitat capacity depends heavily on landscape configuration. Landscapes with excessive fragmentation generally performed poorly. More heterogeneous landscapes tended to support stronger ecological communities.

Indicator Species: Nature's Early Warning System

Perhaps the most practical outcome of the research was identifying indicator species. Indicator species act as biological sensors. By monitoring their presence or absence, we can infer broader ecological conditions without surveying entire communities. Several species emerged as exceptionally reliable indicators:

Indian Marbled Toad
Asian Grass Frog
Common Dotted Garden Skink
Bronze Grass Skink
Small Indian Civet

These species consistently reflected changes in aridity, habitat complexity, fragmentation, and landscape composition. For conservation practitioners, such species provide powerful tools for rapid ecosystem assessment.

What Does This Mean for Conservation?

The study delivers a clear message: Biodiversity conservation in Rajasthan cannot focus solely on protected areas. The landscapes that support wildlife are increasingly agricultural, fragmented, and human dominated.

Conservation success therefore depends on maintaining:

Landscape heterogeneity
Habitat connectivity
Diverse land-cover mosaics
Functional ecological networks

The findings also suggest that agricultural landscapes can support substantial biodiversity when habitat complexity is retained. Field margins, scattered trees, rangelands, wetlands, and natural vegetation patches all contribute to ecological resilience.

Conclusion: The Forward-Looking Ponder

The goal of modern landscape research is to bridge the gaps in our understanding of human-dominated systems. For too long, we have relied on simplified metrics to study environments that are inherently complex and climate-stressed. By integrating spatial structure with trait based approaches, we can see how land-use change—the dominant driver of transformation today—can be steered toward better outcomes.As the primary architects of the Earth's surface, our challenge is to move beyond the factory model of agriculture. Can we redesign our land-use to be truly "multifunctional"? The persistence of the vertebrate communities that engineer our world depends on our ability to arrange our farms not just for production, but as connected geometries of life. The survival of the mosaic is, ultimately, a matter of design.

The Future of Rajasthan's Landscapes

Over years, Rajasthan experienced measurable shifts in land cover, increasing fragmentation, and rapid urban growth. Yet the study also revealed remarkable ecological resilience. Birds, reptiles, mammals, and amphibians continue to find opportunities within these changing landscapes. Their persistence depends not only on the amount of habitat available, but on how that habitat is arranged across the landscape.

The lesson is both scientific and practical:

Landscapes are more than collections of land-cover classes. They are living ecological networks where structure shapes biodiversity, biodiversity influences ecosystem functioning, and indicator species reveal the health of the entire system.

As Rajasthan continues to develop, the challenge will be ensuring that these landscapes remain capable of supporting both human livelihoods and the extraordinary diversity of life that depends on them.

Page updated

Google Sites

Report abuse