Background: The Eastern Ghats, a discontinuous mountain chain extending across the eastern coastal states of India (Andhra Pradesh, Odisha, Tamil Nadu, and Telangana), harbour extraordinary plant biodiversity characterised by high endemism. This mountain system, a component of the Indo-Burma biodiversity hotspot, supports numerous endemic shrub and tree species that are increasingly threatened by accelerating habitat fragmentation driven by agricultural expansion, mining, road construction, encroachment, and invasive species proliferation. Endemic shrub species such as Boswellia ovalifoliolata N.P. Balakar & A.N. Henry, Decalepis hamiltonii Wight & Arn., Rhynchosia beddomei Baker, Andrographis beddomei C.B. Clarke, Tephrosia calophylla Bedd., and Eriolaena lushingtonii (Mast.) Brand, many endemics to specific hill ranges, now persist as fragmented, geographically isolated populations. The consequences of this fragmentation for genetic diversity, gene flow, and long-term population viability remain inadequately quantified at the landscape scale.

Objectives: This study, conducted over a 24-month period (June 2024 to May 2026), aimed to: (1) quantify the extent, severity, and spatial pattern of habitat fragmentation across eight selected Eastern Ghats hill ranges in Andhra Pradesh and Telangana using remote sensing and landscape metrics; (2) assess the population sizes, density, and regeneration status of five focal endemic shrub species across nine fragmented habitat patches; (3) analyse genetic diversity and population structure of the focal species using microsatellite (SSR) markers to quantify gene flow (Nm) and genetic differentiation (F_ST) between isolated patches; (4) evaluate the potential of identified forest corridors and riparian vegetation strips as conduits for pollen and seed dispersal; and (5) develop a prioritised conservation framework incorporating corridor connectivity and population viability analyses.

Methods: Multi-temporal satellite imagery (Sentinel-2A, 10m resolution, 2019 and 2024) was processed using Google Earth Engine to quantify deforestation and fragmentation metrics (patch area, perimeter-area ratio, edge density, Euclidean nearest-neighbour distance, landscape division index) across eight hill range study sites. Field surveys using 10m×10m quadrats sampled 270 plots across nine habitat patches for nine forest fragments during June 2024–May 2025, documenting focal species density, height class distribution, and regeneration indices. Leaf tissue from 486 individuals across the five focal species (from six populations per species) was collected and genotyped using 8–12 microsatellite loci per species. Population genetic parameters (Na, Ne, Ho, He, F_IS, F_ST, gene flow Nm) were computed in GenAlEx v6.503, FSTAT, and STRUCTURE v2.3.4. Least-cost corridor modelling using Circuitscape and ResistanceGA was conducted to identify functional connectivity pathways between populations. Population Viability Analysis (PVA) using Vortex v10.5.3 was run for 100 years under three fragmentation scenarios.

Results: Between 2019 and 2024 (five years), the Eastern Ghats study sites in Andhra Pradesh lost an estimated 4,312 ha of scrub–forest mosaic habitat, with the Nallamalai–Srisailam region recording the highest fragmentation rate (23.4% forest cover reduction). The mean patch area decreased from 186.4 ha to 127.8 ha (reduction of 31.4%) while mean nearest-neighbour distance increased from 2.8 km to 4.6 km. Population genetic analysis revealed alarming erosion of genetic diversity: mean expected heterozygosity (He) = 0.48 ± 0.09 across all focal species–population combinations, with isolated patches showing significantly lower He (0.38 ± 0.06) compared to corridor-connected patches (0.57 ± 0.08). Mean pairwise F_ST among geographically isolated populations was 0.287 ± 0.063, indicating substantial genetic differentiation consistent with severely restricted gene flow (mean Nm = 0.87 migrants per generation). STRUCTURE analysis identified 4–5 distinct genetic clusters corresponding broadly to geographic hill range boundaries. Least-cost corridor modelling identified three priority corridor pathways (Nallamalai–Gundla Brahmeswaram, Seshachalam–Lankamalai, and Papikondalu–Pakhal) that would restore functional connectivity for the focal species. PVA under the baseline (no-corridor) scenario projected 67% extinction probability for populations of Boswellia ovalifoliolata within 50 years.

Conclusion: Habitat fragmentation in the Eastern Ghats has critically compromised the genetic integrity, connectivity, and long-term viability of endemic shrub populations. The significant genetic differentiation among populations, evidence of inbreeding and genetic bottlenecks, and low gene flow estimates collectively indicate that currently fragmented patches are ecologically and genetically isolated functional units on an extinction trajectory. The three identified priority corridors, if conserved and restored through targeted legal protection, agroforestry buffer zones, and community forest management, could restore functional connectivity and reduce 50-year extinction probability from 67% to approximately 22%. These findings provide an evidence-based spatial prioritisation framework for Eastern Ghats endemic shrub conservation under the Biodiversity Conservation Plan of the Andhra Pradesh Forest Department and the National Biodiversity Action Plan 2023–28.