Phylogeography of the Black Kite (Milvus migrans) in Punjab Wetlands: Assessing genetic connectivity and lineage admixture at a migratory crossroads Muhammad Zeshan Haider, Gulnaz Afzal, Hafiz Ishfaq Ahmad, Javed Hussain, Iram Qadeer, Shakeel Ahmed, Laila A AL-Essa, Nisar Ahmad Koka ABSTRACT The Milvus genus presents a taxonomic challenge due to the complex delineation of its evolutionary units. Research on the potentially divergent mitochondrial lineages within Milvus migrans in Pakistan remains sparse, leaving the regional phylogeny largely uncharacterized. This study utilized partial COI gene sequences to evaluate genetic relationships and identify the presence of divergent lineages within Pakistan’s M. migrans populations. Samples collected from three Punjab wetlands (Chashma Barrage, Taunsa Barrage, and Patisar Lake) were integrated with global GenBank reference sequences to evaluate maternal diversity and reconstruct demographic histories. Maternal genetic distances among regional conspecifics were remarkably low, ranging from 0.00% to 1.00%. However, comparisons across more divergent mitochondrial lineages revealed distances reaching 2.00%, particularly between West African and European isolates; notably, these intraspecific values remained distinctly lower than the significant divergence observed with outgroup taxa. Neutrality tests indicated signatures consistent with population expansion or purifying selection. A hierarchical AMOVA confirmed strong continental-scale genetic structuring (65.39% of total variation) contrasting with high gene flow within Punjab’s panmictic local populations. Principal Coordinates Analysis (PCoA) revealed lineage-driven clustering, supporting a homogeneous admixture of three mitochondrial lineages within the Indus Flyway. These results collectively reveal a significant intersection of divergent lineages within Pakistan’s M. migrans populations, highlighting the Punjab region as a critical zone of genetic admixture. While the data indicate high maternal connectivity and extensive haplotype sharing with distant global lineages, they also hint at complex evolutionary histories. These findings underscore the ecological importance of Punjab wetlands as a migratory hub, though the hypothesis of cryptic diversity remains tentative and requires future validation via multi-locus nuclear genomic markers. Introduction The Black Kite ( Milvus migrans ), belonging to the family Accipitridae, is confined to the Old World distributed across Eurasia, Africa, and Australia [ 1 , 2 ]. In Pakistan, Black Kites are common across wetlands and urban centers, where they depend heavily on human waste and carrion [ 3 ]. Despite its broad distribution, Milvus lineages face escalating threats from habitat loss, pollution, agrochemical poisoning [ 4 ], and direct persecution [ 5 , 6 ]. Such pressures necessitate a robust understanding of population genetics to inform effective management [ 7 ]. While the European M. migrans breeding population is well-studied, representing about 10% of the global total, genetic and demographic data for most of its vast Asian and African mitochondrial lineages remain unknown, posing a significant challenge to global conservation efforts [ 8 ]. Phylogeography, the study of spatial patterns in genetic lineages, provides critical insights into evolutionary history and population connectivity across broad ranges [ 9 ]. The taxonomy of Milvus migrans is complex, with five to seven recognized subspecies, including M. m. migrans (European Black Kite), M. m. lineatus (Black-eared Kite), and M. m. govinda (Pariah Kite) [ 2 , 10 ]. These subspecies exhibit overlapping distributions and intergradation zones where hybridization occurs [ 11 ], notably between migrans and lineatus across Western Siberia and Eastern Europe [ 12 ]. Evidence suggests that this contact zone may extend toward the northwestern Himalayas, encompassing Afghanistan and Pakistan, potentially involving all three Eurasian subspecies [ 11 ]. Phenotypic similarities between migratory and sedentary groups further complicate the identification of lineages, highlighting the need for molecular resolution [ 7 ]. In this context, molecular tools, particularly DNA barcoding, have emerged as indispensable for modern biodiversity research and species identification [ 13 ]. This powerful technique involves the rapid and cost-effective analysis of specific mitochondrial DNA regions, notably the cytochrome c oxidase subunit I ( COI ) gene, to identify species and infer evolutionary relationships [ 14 ]. As a protein-coding mitochondrial marker, the COI gene is characterized by rapid evolution and a lack of recombination. These attributes make it ideal for resolving genetic structure, identifying divergent lineages, and reconstructing phylogenetic histories in avian species [ 15 , 16 ]. Pakistan is situated at the confluence of several major avian flyways, most notably serving as the axis for the ‘Indus Flyway’ (International Migratory Bird Route Number 4), also known as the Green Route [ 17 , 18 ]. Stretching from the Karakoram ranges down to the Indus Delta, this corridor is a critical biological artery for over 400 migratory bird species that undertake an exhaustive 4,500 km journey from Siberia and Central Asia [ 19 , 20 ]. The Punjab wetlands situated centrally along the Indus basin are vital migratory bird areas, representing a critical ecological hub for both resident and migratory Milvus species and sub-species, yet their genetic composition remains unexplored [ 21 , 22 ]. Despite the high ecological stakes, critical knowledge gaps regarding the genetic diversity and population structure of raptors along this route hinder evidence-based conservation [ 23 ]. This study, therefore, focused on amplifying and sequencing the mitochondrial COI gene from M. migrans specimens collected across Chashma Barrage, Taunsa Barrage, and Patisar Lake wetlands of Punjab, Pakistan. We integrated our Punjab sequences with a global GenBank dataset to contextualize regional findings within a broader phylogenetic framework. This comparative approach facilitates a deeper analysis of lineage admixture and genetic connectivity at both regional and global scales. By examining local genetic diversity, phylogenetic relationships, and the dynamics of lineage admixture, we aim to elucidate the degree of genetic connectivity characterizing this critical migratory crossroads. This study serves as a foundational mitochondrial assessment, providing a baseline for future genomic inquiries into the potential for cryptic variation within the region. Materials and methods Ethical considerations This study was conducted in rigorous adherence to the ethical protocols for animal research established by the Institutional Bio-Safety Committee (IBSC) and the Department of Zoology (Approval No. 1112/AS&R), The Islamia University of Bahawalpur, Pakistan. Official sanction was granted by the Institutional Animal Ethics Committee (IAEC) via letter No. 454/ORIC, dated December 19, 2024. All research procedures were executed and reported in accordance with the ARRIVE guidelines. To prioritize animal welfare, field procedures were limited to non-invasive handling and blood collection; no protocols involving anesthesia, euthanasia, or animal sacrifice were employed, and no instances of individual animal suffering were observed during the investigation. Study area This study was conducted in three prominent wetlands in the Punjab province of Pakistan: Chashma Barrage (32°25’00.0"N, 71°22’00.0"E), Taunsa Barrage (30°30’46.0"N, 70°50’57.0"E), and Patisar Lake (29°20’41.2″N, 71°56’21.0″E). These locations are crucial stopover places along the Indus Flyway (Green Route), attracting both local and migratory avifauna, including M. migrans populations from Siberia, Russia, and northern regions [ 17 , 22 ]. Chashma Barrage (Site 1) designated wetland encompasses 33,082 hectares. The semi-arid climate, characterized by hot summers (up to 41°C in June) and cold winters (down to 4.5°C in January), provides a diverse environment for wintering migratory birds [ 24 ]. Taunsa Barrage (Site 2) is recognized as an Important Bird Area and a key component of the Indus Flyway. Its rich biodiversity includes globally threatened species and new records documented in our previous research [ 21 ]. Patisar Lake (Site 3) is located within Lal Suhanra National Park, with significant supporting bird species [ 24 ]. Black Kite capture and blood collection During fieldwork Milvus migrans were captured using the bal-chatri method, a widely recognized raptor capture technique by Berger and Mueller [ 25 ]. This trap, which consists of a wire cage adorned with monofilament nooses and containing animal flesh as a lure, is highly effective. The trap size and shape were adjusted according to the target species. A total of 18 M. migrans were captured (6 from each site) across three wetlands; Chashma Barrage, Taunsa Barrage, & Patisar Lake in Punjab, Pakistan. Each specimen was assigned a unique abbreviation for individual identification and locality tracking. Blood samples were collected in sterilized vacutainer tubes utilizing EDTA anticoagulant and stored in an ice box before DNA extraction. Following sample collection, all Black Kites were immediately released back into their natural habitat. DNA extraction Genomic DNA was extracted from the collected blood samples using the Qiagen DNeasy Blood & Tissue Kit, following the standard spin-column protocol for total DNA purification from animal blood [ 26 ]. This protocol is optimized for samples containing nucleated erythrocytes, characteristic of avian species. Prior to extraction, blood samples were carefully thawed if previously frozen, and the manufacturer recommended protocol was strictly followed under sterile conditions. For each sample, 20 µl of Proteinase K was added to a 1.5 ml microcentrifuge tube containing 5–10 µl of anticoagulated blood. The volume was adjusted to 220 µl with phosphate-buffered saline (PBS). In some cases, 4 µl of RNase A (100 mg/ml) was added to digest RNA contamination, ensuring high-quality DNA isolation. The mixture was vortexed thoroughly for homogenization before adding 200 µl of Buffer AL. The solution was then incubated at 56°C for 10 minutes for lysis and protein degradation. Subsequently, 200 µl of ethanol (96–100%) was added to the lysate to promote DNA binding to the spin column silica membrane. The lysate was transferred to a DNeasy Mini spin column placed in a 2 ml collection tube and centrifuged using a Thermo Scientific™ Pico™ 21 Microcentrifuge at 6000 x g for 1 minute. The column underwent a wash sequence with Buffer AW1 followed by Buffer AW2, with flow-through discarded after centrifugation to purify membrane-bound DNA. A final wash was performed before a 3-minute centrifugation at 14,000 x g to dry the membrane and remove residual ethanol. DNA was eluted by applying 200 µl of Buffer AE directly to the membrane, incubating for one minute at room temperature, and then centrifuging at 6000 x g for one minute. The elution step was repeated to maximize DNA yield. PCR amplification and gel electrophoresis The mitochondrial DNA cytochrome c oxidase subunit I ( COI ) gene was selected as a genetic marker due to its widespread use in avian genetic studies and its proven reliability for species identification and preliminary population structure analysis [ 11 ]. DNA amplification of the COI gene segment was performed using polymerase chain reaction (PCR) with the following primers: the forward primer 5’ GGTCAACAAATCATAAAGATATTGG 3’ and the reverse primer 5’ TAAACTTCAGGGTGACCAAAAAATCA 3’ [ 27 ]. PCR amplification was carried out in a thermal cycler (25 μl volume) with the following cycling conditions: initial denaturation at 95°C for 5 minutes, followed by 35 cycles of denaturation at 94°C for 45 seconds, annealing at 55°C for 35 seconds, and extension at 72°C for 45 seconds. A final extension step at 72°C for 5 minutes was included to ensure complete amplification of the target DNA segment. The PCR products were held at 10°C until further processing. The PCR products, approximately 700 bp in size, were separated by 1.5% agarose gel electrophoresis. The resulting DNA fragments were visualized using UV transillumination to confirm successful amplification ( S1 Fig ). PCR products from 18 samples selected randomly from all sampling sites were sent to the Celemics BTSeq™ Institute for sequencing performed by 1st BASE Services, Singapore. The sequencing reactions used the same primers as those used for PCR amplification. Sequence alignment The resulting chromatograms from our sequencing efforts were initially analyzed and edited using Chromas Lite v 2.01 (Technelysium Pty Ltd) to remove any ambiguous bases and refine the sequence data, ensuring high sequence quality. To assess the level of similarity to previously reported sequences and confirm species identification, a Basic Local Alignment Search Tool (BLAST) analysis was conducted against the National Center for Biotechnology Information (NCBI) GenBank database. For multiple sequence alignment, the eight COI partial sequences generated from our Milvus migrans specimens were processed using BioEdit software [ 28 ]. Our newly generated COI partial sequences from the eight Milvus migrans specimens have been deposited under GenBank accession numbers PQ892169 to PQ892176. An additional 36 COI gene sequences of Milvus migrans and closely related Milvus species were retrieved from the NCBI GenBank and Data Portal – BOLD (The Barcode of Life Data System) systems databases. These publicly available sequences originated from diverse geographic locations, including Japan (n = 13), Nigeria (n = 3), Czechia (n = 5), Germany (n = 5), India (n = 4), Thailand (n = 2), Russia (n = 1), and Northern Mariana Islands (n = 1), along with two additional M. migrans sequences from Pakistan (n = 2). In total, 44 COI gene sequences were utilized for the comprehensive analyses ( Table 1 ). 10.1371/journal.pone.0351642.t001 Table 1 Details of Milvus species COI gene sequences used in this study, including newly generated and reference sequences from GenBank. Species Name Origin Accession Reference Milvus migrans Patisar Lake, Pakistan PQ892169 This Study Milvus migrans Taunsa Barrage, Pakistan PQ892170 This Study Milvus migrans Taunsa Barrage, Pakistan PQ892171 This Study Milvus migrans Taunsa Barrage, Pakistan PQ892172 This Study Milvus migrans Chashma Barrage, Pakistan PQ892173 This Study Milvus migrans Taunsa Barrage, Pakistan PQ892174 This Study Milvus migrans Chashma Barrage, Pakistan PQ892175 This Study Milvus migrans Patisar Lake, Pakistan PQ892176 This Study Milvus migrans Japan AB842941 [ 29 ] Milvus migrans Japan AB842942 [ 29 ] Milvus migrans Japan AB842938 [ 29 ] Milvus migrans Japan AB842940 [ 29 ] Milvus migrans Japan AB842939 [ 29 ] Milvus migrans Japan AB842937 [ 29 ] Milvus migrans Pakistan JN801326 [ 30 ] Milvus migrans Japan DBRCJ058 GenBank Submission Milvus migrans Japan DBRCJ059 GenBank Submission Milvus migrans Japan DBRCJ060 GenBank Submission Milvus migrans Nigeria MH536178 GenBank Submission Milvus migrans Nigeria MH536179 GenBank Submission Milvus migrans Nigeria JX160002 [ 31 ] Milvus migrans Czechia KU640398 [ 32 ] Milvus migrans Czechia KU640403 [ 32 ] Milvus migrans Germany KU640407 [ 32 ] Milvus migrans India KP975251 [ 33 ] Milvus migrans India KP975252 [ 33 ] Milvus migrans India KP975253 [ 33 ] Milvus migrans Czechia KU640399 [ 32 ] Milvus migrans Germany KU640400 [ 32 ] Milvus migrans Czechia KU640401 [ 32 ] Milvus migrans Germany KU640402 [ 32 ] Milvus migrans Germany KU640404 [ 32 ] Milvus migrans Germany KU640405 [ 32 ] Milvus migrans Czechia KU640406 [ 32 ] Milvus migrans govinda Thailand MK932893 GenBank Submission Milvus migrans lineatus Thailand MK932894 GenBank Submission Milvus migrans Russia MT773184 [ 27 ] Milvus migrans India OL638139 GenBank Submission Milvus migrans Northern Mariana Islands JQ175385 [ 34 ] Milvus migrans Pakistan JQ175384 [ 34 ] Milvus migrans Japan YIO1026−24 [ 29 ] Milvus migrans Japan YIO1042−24 [ 29 ] Milvus migrans Japan AB843597 [ 29 ] Milvus migrans Japan YIO708−18 [ 29 ] Phylogenetic reconstruction Evolutionary relationships were reconstructed using a robust Bayesian Inference (BI) framework in BEAST v1.10.4 [ 35 ] To adhere to methodological parsimony and avoid uncalibrated temporal inferences, a strict molecular clock was enforced with the clock rate fixed at 1.0. Consequently, branch lengths and the associated scale bar represent genetic divergence in units of substitutions per site. The General Time Reversible (GTR + G + I) model was identified as the optimal substitution model via ModelFinder. A Yule speciation prior was utilized to model branching dynamics [ 36 ]. The Markov Chain Monte Carlo (MCMC) simulation was executed for 10 million generations, with parameters sampled every 1,000 steps. Convergence was rigorously verified in Tracer v1.7, with all parameters achieving an Effective Sample Size (ESS) > 1,500. A Maximum Clade Credibility (MCC) tree was generated using TreeAnnotator after a 10% burn-in, rooted with the White-tailed Eagle ( Haliaeetus albicilla ), and visualized in FigTree v1.4.4. [ 37 ]. Rooting with H. albicilla provided a stable reference for interpreting evolutionary relationships within the ingroup [ 38 ]. Pairwise genetic distances were obtained using the p-distance model in MEGA 11 [ 39 ]. Statistical analysis Genetic diversity indices, including haplotype ( H d ) and nucleotide diversity (π), were computed using DnaSP v6 [ 40 ]. To evaluate the spatial partitioning of genetic variance, we performed a hierarchical Analysis of Molecular Variance (AMOVA) in Arlequin v3.5.2.2 [ 41 ]. Populations were categorized into three hierarchical groups based on biogeographical proximity and phylogenetic affinity: (1) Punjab Local (newly sequenced samples), (2) Asian Regional (reference sequences from Japan, India, Thailand, and Russia), and (3) Extra Regional (divergent lineages from Europe and Africa). Quantitative levels of genetic differentiation were assessed via pairwise F ST values with 10,000 permutations. Demographic history was inferred through Tajima’s D and Fu’s F S neutrality tests. To provide a robust secondary line of evidence for demographic shifts, we generated Mismatch Distributions in DnaSP, comparing observed pairwise differences against expected values under a population growth-decline model. Multivariate analysis, specifically Principal Coordinates Analysis (PCoA), was employed to explore genetic relationships among individuals, and mitochondrial lineage composition was mapped across the study sites. Results Genetic diversity and haplotype connectivity The DnaSP analysis identified a total of 15 distinct haplotypes within the mitochondrial COI dataset (N = 44). These sequences were partitioned into three hierarchical sets: Punjab_Local (n = 8, newly generated), Asian_Regional (n = 23, conspecific references), and Extra_Regional (n = 13, divergent lineages) ( Table 2 ). The Punjab_Local wetland samples exhibited moderate genetic diversity ( H d = 0.464, π = 0.00177), while the Asian Regional group showed slightly lower diversity ( H d = 0.379, π = 0.00095). In contrast, the Extra Regional dataset exhibited the highest levels of variation ( H d = 0.987, π = 0.00937), reflecting its composition of disparate global lineages. 10.1371/journal.pone.0351642.t002 Table 2 Summary of genetic diversity indices for Milvus migrans COI gene sequences. Population/Set N S h H d K π π (JC) Set1 (Punjab_Local) 8 3 3 0.464 0.750 0.00177 0.00177 Set2 (Asian_Regional) 23 2 3 0.379 0.403 0.00095 0.00095 Set3 (Extra_Regional) 13 14 12 0.987 3.974 0.00937 0.00945 Total Data Estimates 44 18 15 0.674 2.413 0.00569 0.00574 N: number of sequences; S: number of segregating sites; h: number of haplotypes; H d : haplotype diversity; K: average number of nucleotide differences; π: nucleotide diversity; π (JC): nucleotide diversity with Jukes-Cantor correction. The TCS haplotype network ( Fig 1 ) visually elucidated the complex maternal relationships among these lineages. A dominant, high-frequency central haplotype (Hap_1) served as a phylogenetic hub, encompassing the majority of local samples from Taunsa Barrage, Patisar Lake, and Chashma Barrage, alongside references from Asian_Regional populations (Japan, Pakistan, India, Thailand). Radiating from this hub were several closely related haplotypes differing by 1–2 mutational steps. Distinct clusters representing Extra_Regional lineages (Nigeria, Europe) were significantly more divergent, forming peripheral nodes in the network. The observed maternal connectivity between Punjab and wider Asian populations, contrasted with the isolation of European/African lineages, underscores the Punjab Wetlands as a significant genetic crossroads for Palearctic Black Kite lineages. 10.1371/journal.pone.0351642.g001 Fig 1 TCS Haplotype network of Milvus migrans. Circle sizes correspond to haplotype frequency. Red segments indicate Punjab_Local samples, green indicates Asian_Regional lineages, and purple represents divergent Extra_Regional lineages. The network illustrates the genetic confluence of local and regional Asian haplotypes and their separation from global lineages. Phylogenetic admixture and lineage relationships The Bayesian phylogenetic tree ( Fig 2 ) distinctly revealed a hierarchical structuring of mitochondrial lineages within Milvus migrans relative to the Haliaeetus albicilla outgroup. The topology highlights the polyphyletic nature of the maternal lineages present within the Punjab local wetlands (indicated in red). The ingroup was broadly divided into three primary clades. The basal clade (Green, posterior probability = 1) predominantly comprised maternal lineages from Europe, specifically Germany (KU640405, KU640404, KU640402, KU640400) and Czechia (KU640398, KU640403, KU640399, KU640401). This clade represents a distinct European M. migrans lineage, likely corresponding to M. m. migrans . Notably, none of the newly generated samples from the Punjab wetlands clustered within this European mitochondrial lineage. In contrast, the remaining Punjab samples were distributed across the more derived Asian and African lineages. A significant secondary clade (Blue, posterior probability = 0.82) grouped our Chashma Barrage sample (PQ892173) with West African lineages from Nigeria (JX160002, MH536179, MH536178). This intermingling suggests a deeper phylogenetic link between certain Indus Flyway populations and Western Palearctic/Afrotropical lineages. The majority of Punjab samples (Patisar Lake PQ892176, Taunsa Barrage PQ892170, PQ892174) were nested within a large, heterogeneous Asian cluster (Yellow/Pink/Purple backgrounds). These local sequences clustered closely with M. m. lineatus and M. m. govinda references from Japan, India, Thailand, and Russia. The polyphyletic distribution of the eight newly sequenced samples across these divergent clades provides robust evidence of the Punjab wetlands as a significant crossroads for multiple mitochondrial lineages. 10.1371/journal.pone.0351642.g002 Fig 2 Bayesian Maximum Clade Credibility (MCC) tree of Milvus migrans based on mitochondrial COI gene sequences. Punjab Wetland sequences are highlighted in red. Values at nodes represent Bayesian posterior probabilities. The tree is rooted with Haliaeetus albicilla . Shaded blocks delineate major clades: Green (European lineage), Blue (African/Chashma admixture), and the large Asian complex (Yellow/Pink/Purple). Scale bar represents genetic divergence in substitutions per site. Genetic distance analysis The maternal genetic distance analysis, utilizing the p-distance model, provided fine-scale insights into the evolutionary relationships among the 44 Milvus migrans COI sequences (detailed in S1 Table ). Overall, the evolutionary divergence patterns among the majority of samples were remarkably low. Most regional sequences exhibited high genetic similarity, with pairwise divergences predominantly ranging from 0.00% to 1.00%. This low distance score indicates that these individuals are closely related and represent conspecific maternal lineages. Specifically, newly generated sequences from Patisar Lake, Taunsa Barrage, and Chashma Barrage (with the exception of PQ892173) showed genetic distances between 0.00% and 1.00% when compared with regional references from Japan, India, Thailand, and Russia. A more pronounced genetic divergence was observed when comparing Indus Flyway populations with extra-regional lineages. Genetic distances between Punjab samples and those from Europe (Czechia and Germany), as well as West Africa (Nigeria), ranged between 1.00% and 2.00%. The European lineage represented by sequence KU640403 exhibited a genetic distance of 2.00% from the Nigerian sequence MH536178. The greatest intraspecific maternal divergence was consistently observed at 2.00%, primarily occurring between Afrotropical and Western Palearctic lineages ( S1 Table ). This significant partitioning underscores the presence of distinct mitochondrial lineages co-occurring within the M. migrans species complex. Hierarchical population structure and admixture The hierarchical AMOVA underscored a profound continental-scale genetic structuring. A significant 65.39% of the total genetic variation was partitioned among the three defined groups ( F ST = 0.653, P < 0.001), while 34.61% was found within populations ( Table 3 ). Analysis of pairwise F ST values delineated a clear hierarchy of genetic connectivity. At the regional scale, we found no significant differentiation between the Punjab_Local and Asian_Regional groups ( F ST = −0.046, P = 0.536), indicating high gene flow and maternal panmixia across the Asian range. This homogeneity contrasted sharply with the genetic distances observed at broader scales. Differentiation between Punjab and Extra_Regional populations was strikingly high ( F ST = 0.703, P < 0.001), confirming the deep divergence of European and African lineages from the Indus Flyway populations. The most profound divisions were observed with geographically isolated lineages from Nigeria and the Northern Mariana Islands ( F ST up to 0.24 in the distance matrix), highlighting the isolating effect of distance on lineage distribution. This pattern of regional connectivity against a backdrop of large-scale structure is summarized in the pairwise heatmap ( Fig 3 ). 10.1371/journal.pone.0351642.t003 Table 3 Hierarchical AMOVA results for Milvus migrans COI sequences based on defined hierarchical groups. Source of Variation df Sum of Squares Variance Components % Variation Fixation Index P -value Among populations 2 701.744 25.296 Va 65.39% F ST = 0.6538* < 0.001 Within populations 41 549.006 13.390 Vb 34.61% Total 43 1250.750 38.687 100% Significance of F -statistic values was tested using a non-parametric permutation approach described in Excoffier, Smouse [ 42 ] with 1,023 permutations: * P < 0.001; df degrees of freedom. 10.1371/journal.pone.0351642.g003 Fig 3 Heatmap of pairwise F ST values. Genetic differentiation among M. migrans populations. Brackets on the axes indicate the hierarchical groups: Punjab_Local, Asian_Regional, and Extra_Regional. Warmer colors indicate higher differentiation; black dots represent non-significant values (P > 0.05). PCoA and lineage composition Multivariate analysis via PCoA clearly separated the influences of evolutionary lineage from geographical origin. The first two axes accounted for 69.9% of the total variation (48.2% and 21.7%, respectively). Visualization by geographic site showed a complete lack of structure among the three Punjab wetlands, with extensive overlap confirming local panmixia ( Fig 4A ). However, when categorized by mitochondrial lineage, three distinct genetic clusters were resolved ( Fig 4B ): an M. m. migrans cluster (negative PCoA1), an intermediate M. m. lineatus cluster, and a well-separated M. milvus cluster (positive PCoA1/PCoA2). The 95% confidence ellipses underscored this pattern, revealing significant differentiation of the M. milvus lineage and substantial overlap between migrans and lineatus , supporting a model of admixture between these two subspecies. 10.1371/journal.pone.0351642.g004 Fig 4 Principal coordinate analysis of COI variation in Pakistani Milvus migrans. (A) PCoA of pairwise genetic distances among individuals from Chashma Barrage, Taunsa Barrage, and Patisar Lake. Points are colored by wetland and shaped by mitochondrial lineage ( M. m. migrans -like, M. m. lineatus -like, M. milvus -like). (B) The same ordination colored by lineage, with 95% confidence ellipses enclosing each lineage cluster. Pakistani individuals largely cluster together, confirming low differentiation among wetlands, but subtle separation of M. milvus -like haplotypes suggests a distinct evolutionary background and potential introgression. The assignment of COI haplotypes to subspecies lineages demonstrated that the Punjab wetland populations constitute a admixed mosaic, with all three mitochondrial lineages M. m. migrans , M. m. lineatus , and M. milvus co-occurring at each site ( Fig 5 ). The M. m. migrans lineage was dominant, representing roughly 67% of individuals, followed by M. m. lineatus , while the M. milvus lineage was consistently present at lower frequencies. Critically, the geographic distribution of these lineages was remarkably homogeneous across the study region. The pie chart visualization ( Fig 5A ) shows only negligible variations in lineage proportions among Chashma Barrage, Taunsa Barrage, and Patisar Lake. This is corroborated by the stacked bar plot ( Fig 5B ), leading to the conclusion that this triple-lineage admixture is a region-wide phenomenon within the Indus Flyway, rather than a site-specific occurrence. 10.1371/journal.pone.0351642.g005 Fig 5 Subspecies lineage composition of Milvus migrans in Punjab wetlands. (A) Schematic map of the three sampling wetlands in Punjab, Pakistan (Chashma Barrage, Taunsa Barrage, Patisar Lake) with pie charts showing the proportion of mitochondrial lineages ( M. m. migrans -like, M. m. lineatus -like, and M. milvus -like) at each site. (B) Stacked barplot summarizing lineage frequencies for each wetland. The coexistence of migrans - and lineatus -like haplotypes, together with a distinct M. milvus -like lineage, illustrates subspecies admixture and supports potential hybridization along the Indus flyway. Demographic inference and neutrality Demographic analyses for the Punjab_Local population yielded a significantly negative Tajima’s D value (−1.701, P = 0.013), which is traditionally consistent with a recent demographic expansion or the signature of purifying selection acting on the mitochondrial genome. However, the Mismatch Distribution for the overall dataset ( Fig 6 ) exhibited a distinctly multimodal pattern, diverging from the unimodal bell curve expected under a simple population expansion. This complexity, supported by a Harpending’s Raggedness index ( r = 0.0524) and R 2 statistic (0.0604), showed that the observed genetic signatures likely reflect the confluence of multiple, divergent lineages with independent demographic histories (the “crossroads” effect) rather than a single, cohesive expansion event. 10.1371/journal.pone.0351642.g006 Fig 6 Mismatch distribution of Milvus migrans COI sequences. The observed frequency of pairwise differences (red line) compared to the expected population growth-decline model (dashed green line). The multimodal pattern shows the intersection of multiple divergent genetic lineages within the dataset. Discussion This research aimed to provide the first integrative genetic assessment of Black Kite populations in Punjab wetlands (Chashma Barrage, Taunsa Barrage and Patisar Lake), clarifying their evolutionary lineages and potential interactions with other Milvus taxa. Our analysis identified substantial mitochondrial variability across both regional and global scales, characterized by a moderate degree of haplotype diversity and significant negative neutrality statistics. While the observed negative Tajima’s D in local populations is consistent with demographic expansion following historical bottlenecks [ 43 , 44 ]. We acknowledge that mitochondrial COI data are intrinsically susceptible to purifying selection and selective sweeps, which can manifest as similar neutrality signatures. This interpretation is further nuanced by the distinctly multimodal nature of our mismatch distribution; rather than reflecting a singular, cohesive expansion event, this pattern likely denotes a confluence of disparate lineages (the ‘crossroads’ effect). Consequently, we interpret these findings as reflective of complex maternal evolutionary dynamics, the definitive resolution of which remains contingent upon the integration of multi-locus nuclear data to distinguish between regional demographic growth and natural selection on the mitochondrial genome. At the local scale, the absence of significant genetic differentiation among the three Punjab wetlands suggests a highly connected and panmictic population. This regional homogeneity is likely facilitated by the Black Kite inherent mobility [ 45 , 46 ], a pattern consistent with other wide-ranging raptors where extensive dispersal minimizes regional divergence [ 47 ]. However, this local panmixia starkly contrasts with the robust genetic structuring observed at broader, continental scales. Our study results revealed considerable barriers to gene flow beyond regional boundaries. This strong partitioning of genetic variation suggests that while populations within a connected landscape like the Punjab wetlands may interbreed freely, larger geographical distances and barriers effectively limit gene exchange between continental populations. The phylogenetic reconstruction underscores the Punjab region role as a critical biogeographical hub. The most striking finding is the polyphyletic distribution of the local maternal lineages; rather than forming a single, isolated population, the Punjab samples are dispersed among global clades. Specifically, the nesting of the majority of local sequences within the Asian lineage cluster confirms a strong genetic affinity with eastern Palearctic populations. However, the consistent grouping of the Chashma Barrage sample (PQ892173) with divergent lineages from the Afrotropical/Western Palearctic region illustrates a significant aspect of mitochondrial diversity that has been previously overlooked in South Asia. This aligns with previous studies that have identified overlapping ranges and hybridization between M. m. migrans and M. m. lineatus , particularly in areas like Central Asia and the Himalayas [ 2 , 10 ]. The challenges in distinguishing these subspecies genetically, especially in regions of overlap where migrants and residents co-occur, have been noted in prior research [ 7 , 32 ]. The genetic distance between M. migrans and M. milvus is generally small, reflecting their divergence during the Pleistocene, but they are typically considered reproductively isolated species, albeit with incomplete isolation evident through occasional hybridization [ 32 , 48 ]. Documented hybridization between M. milvus and M. migrans in Central Europe [ 49 – 51 ], confirms that interbreeding occurs among these taxa, albeit infrequently in raptors [ 52 ]. In our study, this intermingling of Pakistani samples with distinct mitochondrial lineages suggests that the Indus Flyway serves as a zone of secondary contact or a refugial crossroads for divergent migratory routes. The 0.00% to 2.0% genetic distances observed are characteristic of intraspecific lineage admixture rather than definitive speciation. These results provide a foundational mitochondrial framework that necessitates future validation with multilocus nuclear markers to resolve the biparental dynamics of hybridization and introgression at this ecological intersection [ 14 ]. Our study contributes significantly to filling a critical knowledge gap in the genetic characterization of M. migrans in South Asia. While previous phylogenetic studies on Black Kites were also limited by sample size, geographic representation and utilized published nucleotide sequences [ 2 , 10 , 32 , 48 , 53 – 55 ]. Our study highlights the region as a vital crossroads for Milvus lineages, underscoring the necessity for conservation strategies that account for genetic connectivity and potential admixture. Future research should prioritize expanded geographical sampling across Pakistan and incorporate multiple nuclear markers to resolve the evolutionary history of Milvus species in this critical part of their range. Conclusion Our comprehensive genetic analysis of M. migrans populations in the Punjab wetlands has revealed notable genetic variability, identifying 15 distinct haplotypes. This analysis underscores the complexity and diversity of maternal lineages within the species. The phylogenetic clustering of our Punjab Wetland isolates among diverse global lineages suggests that this region serves as a vital confluence for Milvus migrans populations. Rather than confirming cryptic speciation, our data point toward a shared yet intricate evolutionary history shaped by high connectivity and the convergence of distinct Palearctic lineages. Supporting information S1 Table Pairwise genetic distances ( p -distance) among Milvus migrans mitochondrial COI gene sequences. The lower triangular part displays genetic distances in percentages; the upper triangular part shows the associated standard errors. (PDF) S1 Fig Original uncropped and unadjusted gel electrophoresis images of mitochondrial COI gene PCR products. Electrophoretic separation was conducted on a 1.5% agarose gel and visualized via UV transillumination. Lane M denotes a 100 bp DNA ladder (ranging from 100 bp to 1500 bp). The observed bands at approximately 700 bp confirm the successful amplification of the target COI gene segment for the Punjab wetland samples (labeled P1–P3 and Q2–Q6). (PDF)