Recent historical bottlenecks and restricted gene flow in one of the last remaining stronghold populations of the southern black-throated finch (Poephila cincta cincta)

Recent historical bottlenecks and restricted gene flow in one of the last remaining stronghold populations of the southern black-throated finch (Poephila cincta cincta) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Recent historical bottlenecks and restricted gene flow in one of the last remaining stronghold populations of the southern black-throated finch (Poephila cincta cincta) Skye Davis, Adam J Stow, Jemma McCrossin, Wilbur Ashley, John M van Osta, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6780931/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 6 You are reading this latest preprint version Abstract Habitat fragmentation is a key driver of reduced genetic connectivity and loss of genetic variation among populations, elevating the risk of inbreeding depression and reduced adaptive potential. The endemic, nationally endangered southern black-throated finch ( Poephila cincta cincta ) has experienced severe range contractions since the rise of pastoralism. Using a panel of genome-wide single nucleotide polymorphisms, we characterised spatial genetic structure for a remaining stronghold population in the Desert Uplands Bioregion of Queensland, Australia. We mapped effective migration surfaces and tested for isolation by resistance to identify potential barriers to gene flow, estimated contemporary effective population sizes and reconstructed the demographic history of this population. We found evidence of restricted gene flow between localities only 16 km apart and strong isolation by geographic distance. Landscape resistance modelling identified areas of suitable woodland habitat that facilitated effective dispersal. More restricted gene flow in the southern range of this population is likely influenced by the fragmentation of suitable vegetation communities. Contemporary effective population sizes were near or below 1000, and we detected two historical population bottlenecks (> 50% decline) occurring around 60–100 and 700 years ago. Given recent evidence that the Desert Uplands population is genetically isolated from the only other stronghold population in Townsville, the results of this study suggest future losses of genetic diversity and adaptive potential may continue without effective management. To improve the long-term persistence of southern-black throated finch across their range, prioritising the conservation and restoration of habitat that promotes genetic connectivity is essential. Effective population size kinship landscape resistance modelling functional connectivity historical demography conservation genetics Figures Figure 1 Figure 2 Figure 3 Introduction Ongoing species declines are occurring despite global conservation efforts and environmental protection laws. Biodiversity is being lost at an unprecedented scale, and the current extinction rate is higher than the background rate for many taxonomic groups, particularly birds, amphibians and mammals (Brondízio et al., 2019 ). As the fundamental source of biodiversity, conserving genetic diversity in wild populations is vital for supporting the long-term persistence of species (Allendorf & Luikart, 2009 ; Lowe & Allendorf, 2010 ; Raffard et al., 2019 ). Reduced gene flow, for example due to habitat fragmentation, can lead to a loss of genetic diversity and an increased risk of inbreeding depression, reducing adaptive potential and elevating extinction risk (Frankham, 2005 ; Frankham et al., 2017 ). More than 10% of global genetic diversity may have already been lost based on the relationship between decreasing habitat and the loss of genome-wide diversity (Exposito-Alonso et al., 2022 ), and a 19–66% decline in genetic diversity is forecasted without intervention (Hoban, Bruford, et al., 2021 ). Consequently, molecular approaches have been widely adopted in wildlife management, particularly with the rise of high-throughput genotyping technologies and whole genome sequencing (Hoban et al., 2022 ; Holderegger et al., 2019 ; Pierson et al., 2016 ). Despite these advances, there is a severe lag in incorporating genomic data into management plans and policies, and challenges remain with how genetic data is utilized in an effective and timely manner for conservation (Hoban, Campbell, et al., 2021 ; Walters & Schwartz, 2020 ). The black-throated finch ( Poephila cincta , hereafter referred to as BTF) is a granivorous (seed-eating) finch endemic to Australia with two recognised subspecies, a northern ( P. c. atropygialis , BTFN) and southern form ( P. c. cincta , BTFS) (Higgins, 2006 ). BTF have declined since the early 1900’s with the rise of pastoralism, and their range has contracted by more than 80% since the late 1900’s (BTFRT, 2007 ; Franklin, 1999 ; Reside et al., 2019 ). The southern subspecies was formerly distributed from north-eastern NSW up to the Atherton Tablelands (Baldwin, 1976 ), but its range is now highly fragmented and restricted to north-eastern QLD (Schodde & Mason, 1999 ). Two strongholds remain; the Townsville Coastal Plain and the Desert Uplands Bioregion (Galilee Basin), where the largest remaining population is found (Grice et al., 2023 ; Vanderduys et al., 2016 ). Although BTF is listed as ‘Least Concern’ on the IUCN Red List of Threatened Species (BirdLife International, 2024), BTFS was nationally listed in 2005 as “Endangered” (Environment Protection and Biodiversity Conservation Act 1999), and state-listed in 2016 as “Presumed Extinct” in NSW (Threatened Species Conservation Act 1995). Yet despite increased environmental regulation since BTFS was first listed under these environmental protection laws, the southern subspecies has continued to decline in recent years (Reside et al., 2019 ). Conserving the last two strongholds is crucial for ensuring the long-term persistence of BTFS. The major threats facing BTFS are habitat loss and degradation, particularly in the south of its range where agriculture has been more intensive (Reside et al., 2019 ). BTF have strict habitat requirements, occupying open grassy woodlands and forests with regular access to drinking water (Higgins, 2006 ; Zann, 1976 ). Based on flock sizes and radio tracking data, BTFS generally prefer habitats close to permanent water sources, with low tree and shrub cover and a higher cover and richness of native grasses (higher vegetation density within the 0–20 cm height class), avoiding heavily grazed areas with low grass diversity (Rechetelo, 2015 ; Rechetelo et al., 2016 ; van Osta et al., 2024 ). The decline of BTFS has been primarily driven by agriculture, clearing of grassy woodlands, habitat degradation by livestock grazing and the spread of invasive species (e.g., exotic grasses) (BTFRT, 2007 , 2020 ; Forshaw & Shephard, 2012 ; Garnett et al., 2011 ; Woinarski & Catterall, 2004 ). More than half of the remaining habitat for BTFS is threatened by mining, particularly affecting the larger resident population in the Desert Uplands Bioregion (Vanderduys et al., 2016 ). The Townsville population is predominantly threatened by urbanization and land development (BTFRT, 2020 ; Mula Laguna et al., 2019 ). Flock sizes have decreased in both populations (Mula Laguna et al., 2019 ), but are larger in the Desert Uplands Bioregion, likely due to a higher percentage of remnant vegetation and intact BTF habitat compared to Townsville (Reside et al., 2017 ). BTFS are also considered sedentary, maintaining small home range sizes and exhibiting limited post-breeding dispersal, which may exacerbate their risk of location extinction (Garnett et al., 2011 ; Higgins, 2006 ; Rechetelo et al., 2016 ; van Osta et al., 2024 ). Recovery efforts over the past two decades have aimed at improving ecological knowledge available for BTFS to inform conservation status and guide management strategies (BTFRT, 2007 , 2020 ). Yet conservation remains hampered by limited information on its ecology and a lack of long-term monitoring in the face of ongoing habitat loss (Mula Laguna et al., 2019 ; Reside et al., 2019 ). Information on the movement ecology of BTFS has largely been based on data from the Townsville population (Mula Laguna et al., 2019 ), although more recently, radio tracking was used to investigate habitat use, home range and daily movements in the Desert Uplands population (van Osta et al., 2024 ). There have also been relatively few genetic studies on BTFS. A recent whole-genome sequencing (WGS) study supports the distinction of two subspecies of BTF, suggesting the Einasleigh Uplands restricts gene flow between BTFN and BTFS with no evidence of recent admixture (Hooper et al., 2025 ). A fine-scale study of the Townsville BTFS population, based on microsatellite markers, found genetic structuring over relatively short distances (10–20 km), with differences in elevation and vegetation structure, along with the Ross River Dam, contributing to altered gene flow (Tang, 2016 ; Tang et al., 2017 ). WGS data also supports strong spatial genetic structure within the Townsville population, as well as genetic differentiation between the Townsville and Desert Uplands (Galilee Basin) populations (Hooper et al., 2025 ). However, further research is needed to examine genetic structure and diversity within the larger, resident Desert Uplands population of southern black-throated finch and assess potential genetic consequences of recent population declines. Genetic data can help improve population viability models for BTFS and assess the long-term persistence of this subspecies in the face of environmental change. Here, we use a panel of genome-wide single nucleotide polymorphisms (SNPs) to investigate the functional connectivity and genetic health of southern black-throated finches ( P. c. cincta ) within the Desert Uplands Bioregion population centred around the Carmichael Coal Mine (CCM). We aim to 1) assess spatial genetic structure, diversity and site fidelity; 2) identify potential biogeographic barriers to dispersal with landscape resistance modelling; and 3) estimate effective population size and infer demographic history. Materials and Methods Sample collection Blood samples were collected from black-throated finches on the Moray Downs property (Bravus Mining and Resources) in the Desert Uplands Bioregion of Queensland, Australia, between 2021 and 2023 (Williams et al., in prep ). Samples were collected from three localities (denoted North, Central and South regions, see Fig. 1 ). Sequencing, SNP calling and filtering 158 BTFS blood samples were sent to Diversity Arrays Technology ( http://www.diversityarrays.com ) for DArTseq genotyping-by-sequencing (GBS) on an Illumina platform (Kilian et al., 2012 ; Sansaloni et al., 2011 ). Library construction was carried out using the DArTseq complexity reduction approach. Genomic DNA was digested with PstI and Sp4 restriction enzymes, and barcoded adapter-ligated fragments were amplified with PCR. Libraries were sequenced on an Illumina HiSeq®2500 platform following manufacturer instructions, and samples were demultiplexed according to their ligated barcode. Raw reads were processed to remove barcodes using FASTX-Toolkit v 0.11.8 ( http://hannonlab.cshl.edu/fastx_toolkit/ ), Illumina adapters were removed using AdapterRemoval v 2.3.1 (Schubert et al., 2016 ), and reads with Phred Q-scores > 25 were retained. Quality was checked with FastQC v 0.11.7 ( http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ ). For analysis of neutral genetic structure, reads were checked for contamination against the Bacteria NCBI database and processed using DArT’s proprietary pipeline DArTSoft14™ (Diversity Arrays Technology P/L). SNPs were called de novo by first calling sequencing clusters from pooled samples and then individual samples, removing monomorphic sequence clusters. SNPs were filtered with the R package dartR v 2.7.2 (Gruber et al., 2018 ) by applying the following filters: average reproducibility > 0.98; mean read depth \(\:0<rd 0.95; sample call rate > 0.90; minor allele count (MAC) ≥ 3; retaining one SNP per RAD tag at random; removal of loci that deviated from HWE within populations using the exact tests of Wigginton et al. ( 2005 ) with Bonferroni corrections; and removal of duplicate samples with SNPRelate v 1.34.1 (Zheng & Zheng, 2013 ) where kinship coefficient exceeded 0.45. To generate a set of putatively neutral SNPs, any loci detected as outliers by three independent methods were removed, OutFLANK v 0.2 (Whitlock & Lotterhos, 2015 ), BayeScan v 2.1 (Beaumont & Balding, 2004 ; Foll & Gaggiotti, 2008 ) and Arlequin v 3.5 (Excoffier & Lischer, 2010 ). Close relatives were identified and removed to obtain an ‘unrelated’ dataset, using the functions ‘pcrelate’ and ‘pcairPartition’ from GENESIS v 2.32.0 to update KING kinship coefficients in the presence of population stratification (Conomos et al., 2016 ). The ‘unrelated’ dataset was used for most downstream analyses of genetic structure, owing to the substantial impact close relatives had on genetic structure (Figure S1 ). The ‘related’ dataset (all samples) was used for identifying recent migrants between populations. For analysis of historical demography, raw reads were further processed to retain reads with Q-scores > 33 and were mapped to a reference genome using Bowtie2 v 2.4.5, Picard v 2.26.2 and SAMtools v 1.3.1 (Li et al., 2009 ). As no reference genome was available for BTF, we used the RefSeq zebra finch ( Taeniopygia guttata ) genome assembly ‘bTaeGut1.4.pri’ (GCF_003957565.2) from NCBI GenBank. Sex inference The sex of each bird was inferred using a method adapted from seGMM (Liu et al., 2022 ) and seXY (Qian et al., 2017 ). First, unfiltered SNPs were mapped against the Z and W chromosomes of the zebra finch ( Taeniopygia guttata , RefSeq assembly GCF_003957565.2) using the function ‘gl.blast’ in dartR . Mapped sequences were filtered to retain those with percent identity > 80%, E-value 0.98, and removing loci located in pseudoautosomal regions or mapping to both Z and W chromosomes. Three sex-associated metrics were estimated per individual: W missing rate, Z/W ratio and Z heterozygosity. Sex was inferred with the model-based clustering function ‘Mclust’ from mclust v 6.0.0 (Scrucca et al., 2016 ), with the expectation-maximisation (EM) algorithm to cluster samples into two groups (male or female) (Figure S2). Population genetic analyses Genetic structure was analysed using four individual-based approaches: 1) a principal coordinates analysis (PCoA) with ade4 v 1.7–22 using Euclidian genotypic distance between samples; 2) a discriminant analysis of principal components (DAPC) with adegenet v 2.1.10 (Jombart & Ahmed, 2011 ); 3) SNMF clustering with LEA v 3.14.0; and 4) the Bayesian clustering approach implemented in STRUCTURE v 2.3 (Jombart, 2008 ) via dartR. For the DAPC, we compared results using a priori groups and de-novo clustering, as these approaches can impact inference of genetic structure when population differentiation is low (Miller et al., 2020 ). The optimal number of PC’s was chosen using the function ‘optim.a.score’ with 1000 simulations. STRUCTURE settings included no prior site information and used the admixture model with correlated allele frequencies (Falush et al., 2003 ), with a burn-in/MCMC setting 10,000/100,000 and 5 independent runs for each value of 𝐾 = 1 to 𝐾 = 6. The most likely value of 𝐾 was identified using the Evanno method (Evanno et al., 2005 ), and CLUMPP v 1.1.2 to consolidate results from multiple runs (Jakobsson & Rosenberg, 2007 ). To assess recent migration and identify individuals with mixed ancestry, STRUCTURE was repeated for the ‘related’ dataset, using regions (north, central, south) as prior site information, 𝐾 = 3, ‘gensback’ = 3 (i.e., back to great-grandparents) and migration prior = 0.05. Spatial genetic structure was examined using spatial PCA (sPCA) and spatial autocorrelation analysis (SAA). sPCA detects cryptic spatial structure by decomposing genetic variation into variance and spatial autocorrelation (Moran’s index, I ) with adegenet . sPCA was run using an inverse distance matrix for weighting connectivity in the neighbour network (type 7), and significant global (high variance and positive I ) and local structures (high variance and negative I ) were detected with permutation tests (999 replications). SAA was performed with dartR , following Smouse and Peakall ( 1999 ), using pairwise genotypic correlations ( r ) based on Euclidian genotypic distance, uneven distance classes to optimize sample sizes, 9999 bootstraps to compute 95% C.I’s around mean r within distance classes and 9999 permutations to test the null hypothesis of no spatial structure (i.e., random mating). Sex-biased dispersal was assessed using kinship permutation tests and the sex-biased dispersal tests described in Goudet et al. ( 2002 ). Mean kinship was compared between sexes, localities and regions using the KING-derived kinship coefficients with GENESIS, accounting for genetic structure. Two-sided permutation tests owing to the non-independence of pairwise kinship data, by reshuffling group labels among samples without replacement (999 permutations). Goudet’s sex-biased dispersal tests were implemented with the function ‘sexbias.test’ in the R package Hierfstat (Goudet, 2005 ), using two-sided permutation tests to assess whether four statistics were significantly different between males and females; \(\:{F}_{ST}\) , \(\:{F}_{IS}\) , mean corrected assignment index ( \(\:mAIc\) ) and variance ( \(\:vAIc\) ). Sex was reshuffled for each permutation, keeping sex ratios and individual locations (site or region) constant. Global \(\:{F}_{ST}\) and \(\:{F}_{IS}\) , and pairwise \(\:{F}_{ST}\) between localities or regions were estimated using diveRsity v 1.9.90 (Keenan et al., 2013 ) and dartR (95% C.I.’s computed with 999 bootstraps). Genetic diversity metrics were calculated per region (North, Central, South) with diversity , including allelic richness ( \(\:Ar\) ), observed heterozygosity ( \(\:{H}_{o}\) ), unbiased expected heterozygosity ( \(\:{H}_{e}\) ), and inbreeding coefficient ( \(\:{F}_{IS}\) ). Landscape resistance modelling To identify potential barriers to gene flow between BTFS populations in the Desert Uplands Bioregion, we first mapped estimated effective migration surfaces (EEMS) across the study area using a 15km buffer around all sampling locations. EEMS infers deviation from continuous gene flow across the landscape to identify areas of higher or lower than average historic migration (Petkova et al., 2016 ). In brief, EEMS creates a triangular grid (based on the number of subpopulations, or demes) overlayed on the study area, and estimates the expected genetic dissimilarity between pairs of samples assuming a stepping-stone model of migration between demes. Pairwise resistance distances between samples are calculated using a circuit theory-based approach and the estimated effective migration rates are interpolated across geographic space. As the number of demes chosen affects the grid resolution, we tested a range of deme sizes from n = 50 to n = 500. EEMS was run with default parameters (MCMC iterations = 8,000,000, burn-in = 1,000,000, thinning = 9999), with three independent Markov Chain Monte Carlo (MCMC) chains to assess convergence and average results across runs. Effective migration rates were mapped across the landscape using the R package reemsplot2 v 0.1.0 (Petkova, 2024 ). Functional connectivity across the landscape was further explored using the R package ResistanceGA v 4.2.10 (Peterman, 2018 ) to test for isolation by resistance (IBR). ResistanceGA first optimizes individual landscape resistance surfaces by fitting a linear mixed effects model with a maximum likelihood population effects parameterization (MPLE) to the data, with pairwise genetic distance as the response variable and pairwise effective distance as the predictor variable. ResistanceGA then uses a circuit theory-based approach to measure current flow between individuals across each resistance surface, using the package Circuitscape v 5.13.1 (Hall et al., 2021 ) within Julia v 1.10.0 (Bezanson et al., 2017 ). We assessed the impact of seven landscape variables on inter-individual genetic distance (calculated with ‘gl.dist.ind’ in dartR with method = Euclidian). Previous studies indicate that vegetation density and structure, as well as distance to permanent water sources, influence BTFS dispersal and flock sizes (Rechetelo, 2015 ; Rechetelo et al., 2016 ; van Osta et al., 2024 ), and influence genetic structure in the Townsville population (Tang, 2016 ; Tang et al., 2017 ). In this study, we tested three categorical and four continuous landscape variables that captured vegetation structure, habitat suitability and water availability (Tables 1 , S1, Figures S3-4). Details on the preparation of these landscape raster layers are provided in the Electronic Supplementary Material. Roads and tracks were excluded from the analysis because the study area is intersected by only one public road which receives a low volume of traffic, and BTFS have been observed crossing this road as well as nesting and foraging along the edge (Brad Dreis pers. comm. ). First, each landscape variable was optimized independently using the functions ‘GA.prep’, ‘JL.prep’ and ‘SS.optim’ in ResistanceGA . For each continuous resistance surface, we tested all eight data transformations available in ResistanceGA to choose the optimal transformation. The function ‘JL.prep’ was used to generate pairwise resistance matrices via Circuitscape in Julia. ‘SS.optim’ carried out resistance surface optimization for each layer, with two independent runs (random seed set for each run), as well as for a geographic distance only model and a null model (intercept only). We used the Akaike information criteria corrected for small sample sizes ( \(\:{AIC}_{c}\) ) from the fitted MLPE model to assess the support of each resistance surface model (Beninde et al., 2024 ; Shirk et al., 2017 ). ‘Pseudo’ bootstrapping was carried out for the best run (lowest \(\:{AIC}_{c}\) ) for each surface using the function ‘Resist.boot’. Correlations between landscape layers (both pre- and post-optimization) were calculated using Pearson correlation coefficients and the function ‘layerStats’ from the R package raster v 3.6–26. To test whether models based on only one or multiple landscape variables were better supported, and to assess functional connectivity over the entire landscape, we created optimized multi-feature resistance surfaces (OMFRS) using an iterative approach. We only included models that significantly impacted functional connectivity more than geographic distance alone, i.e., those with a bootstrapped \(\:{AIC}_{c}\) score 0.6, retaining the surface with the smallest \(\:{AIC}_{c}\) ). This function iteratively adds optimized single resistance surfaces to a combined model, starting with the surface with the lowest \(\:{AIC}_{c}\) , adding the surface with the next lowest \(\:{AIC}_{c}\) , and repeating the optimization and bootstrapping procedure. Each single surface layer added sequentially to the combined model was only retained if the bootstrapped \(\:{AIC}_{c}\) of the new combined model was lower than that of the prior model. Lastly, we mapped the least-cost resistance pathways between samples with Circuitscape , for each optimized resistance surface (best run based on lowest \(\:{AIC}_{c}\) ). Contemporary effective population size Contemporary \(\:{N}_{e}\) was estimated using the bias-corrected linkage disequilibrium (LD) method with NeEstimator v 2.1 (Waples & Do, 2008 ), with \(\:{P}_{crit}\) = 0.02 and calculating bias-corrected jack-knife 95% confidence intervals (Waples and Do 2010 ). Given the presence of overlapping generations for BTF, this raw estimate was interpreted as the effective number of breeders in the parent generation which produced the cohorts sampled ( \(\:Raw{\widehat{N}}_{b}\) ). \(\:Raw{\widehat{N}}_{b}\) was adjusted to account for this bias and converted to \(\:{N}_{e}\) using the equations in Table S2, and two life history traits (age at sexual maturity (α) and adult lifespan ( \(\:AL\) )). To account for the limited ecological knowledge available for BTF, we converted \(\:{N}_{e}\) using α = 0.5 or 1 year, and \(\:AL\) = 4 to 8 years (Shephard, 1989 ). No adjustments were made for physical linkage as we assumed the filtered SNP data were physically unlinked (retaining one SNP per RAD tag). As LD is built up over generations, we interpreted \(\:{N}_{e}\) as the short-term harmonic mean across a few recent generations. We also estimated \(\:{N}_{e}\) separately for the North-Central (n = 87) and South (n = 20) groups. To account for differences in sample size, we generated 100 bootstrap replicates for the North-Central group by randomly subsampling 20 individuals each time (with replacement) and recalculating \(\:{N}_{e}\) . Demographic history Demographic history was inferred using two site frequency spectrum (SFS) based approaches; Stairway Plot v 2.0 (Liu & Fu, 2015 ) and δaδi v 2.3 (Gutenkunst et al., 2009 ). For both approaches, 1-dimensional SFS’s were generated with ANGSD v 0.940, using the genome-mapped BAM files and applying the following filters: minimum base quality score > 20, minimum mapping quality > 20, individual read depth > 2 (in at least 50% of samples) and < 1000, doSaf = 1 to calculate allele frequency likelihoods based on individual genotype likelihoods (GL) obtained using SAMtools (GL = 1). Folded SFS’s were generated as we lacked information on ancestral allelic states. Stairway Plot was run with default settings, with sequence length ( \(\:L\) ) calculated as the total number of variant and invariant sites after filtering. Demographic inference with δaδi was performed with dadi-cli (Huang et al., 2023 ). Five 1D models were tested, described in Table S3. Parameters for each model were optimized using the function ‘InferDM’ with five independent runs each with 5000 optimizations and the --global-optimization flag, until convergence was reached (‘BestFits’). As loci were considered ‘linked’, models were compared using model scores (best model yields a score of 1) and Akaike weights ( \(\:{W}_{AIC}\) ) (evaluates the relative probability of each model, as outlined in Rougeux et al. ( 2017 ) (equations in Table S4). 95% confidence intervals around parameter estimates were obtained with 100 bootstrapped SFS’s generated with easySFS (with replacement) and the ‘StatDM’ function in dadi-cli . Parameters for the best-fit model were converted into measures of \(\:{N}_{e}\) and time in years, using the equations listed in Table S5. As the germline mutation rate (per site per generation, \(\:{\mu\:}_{G}\) ) is unknown for BTF, conversions used the estimate for zebra finch, \(\:{\mu\:}_{G}\) = 5.85 \(\:\times\:\) 10 −9 (Bergeron et al., 2023 ), and a generation length ( \(\:g\) ) of 2 years (Bird et al., 2020 ). To account for the limited ecological information available for BTF, we also used a longer generation length of 3.5 years (Garnett et al., 2011 ). Results After filtering, we retained 14,788 neutral SNPs (107 samples) for analyses of neutral genetic structure and diversity, excluding highly related individuals (full-sibs and half-sibs, n = 48) (Table S6). Sex inference identified 70 females and 47 males in the ‘unrelated’ dataset (81 F and 74 M total). For analyses of historical demography, we retained 4,438,791 sites (including 80,382 variant sites) after filtering for Stairway and δaδi analyses (all samples pooled). Evidence of fine-scale genetic structure Using the ‘unrelated’ dataset, global \(\:{F}_{ST}\) indicated weak but significant genetic structure ( \(\:{F}_{ST}\) = 0.011). SNMF clustering with LEA provided support for a single ancestral population, with \(\:K\) = 1 yielding the lowest cross-entropy. The DAPC, STRUCTURE and PCoA analyses all suggested \(\:K\) > 1, and consistently clustered most South samples separately from Central and North samples (Figs. 1 , S5-6). Pairwise \(\:{F}_{ST}\) was significantly different from 0 between all regions but was highest between North and South ( \(\:{F}_{ST}\) = 0.018, p < 0.001), followed by Central and South ( \(\:{F}_{ST}\) = 0.0136, p < 0.001), and North and Central ( \(\:{F}_{ST}\) = 0.0057, p 0.05) to 0.025 p < 0.001) and was highest between localities in the South (CMB or LFT) and all other localities in the Central and North regions (p < 0.001 for all comparisons) (Figure S7). Measures of genetic diversity were similar between regions, although \(\:Ar\) and \(\:{F}_{IS}\) were both significantly higher in the Central sub-population (Table S7). Global \(\:{F}_{IS}\) was 0.160. Based on assignment tests in STRUCTURE with prior population information, most samples from the Central region (97.9%) were correctly assigned to this region with high probability (P > 0.9), and none were identified as migrants from another region (Table S8). Most northern samples (81.5%) also assigned to the central region, with only 7.4% (n = 2) assigning to the north with no mixed ancestry and 11.1% (n = 3) having recent ancestry in the central region (back to grandparents) (P > 0.9, p < 0.001). In the southern region, only 22.6% of samples were correctly assigned to this region with no mixed ancestry, 29.0% were identified as migrants from the central region (n = 9) and 16.1% (n = 5) had recent ancestry in the central region (grandparents) (P > 0.9, p k > 0.1875) across 48 samples, and 61 pairs of second-degree relatives (half-sibs or grandparent-offspring, 0.1875 > k > 0.09375) across 74 samples, following the midpoint cut offs in Iacchei et al. ( 2013 ). First-degree relatives were detected across years and wet/dry seasons and first- or second-degree relatives were always found in the same region (Table S9). Kinship permutations tests indicated significant fidelity, at both local and regional scales, with mean within-site and within-region kinship being significantly higher than between-site and between-region kinship, respectively (p < 0.001) (Figure S8, Table S10). The sPCA indicated significant global structure (p < 0.01), i.e., high variance and positive spatial autocorrelation (Figure S9). SAA revealed a strong signal of isolation by distance (IBD), with mean genotypic correlation ‘ r ’ among samples being significantly different from 0 for all distance class bins (Figs. 1 , S10). Mean ‘ r ’ was significantly higher than expected under conditions of random mating for the 2.5 km distance class (i.e., within-site comparisons) and significantly lower than expected for distance classes ≥ 30 km (Figs. 1 , S10). This pattern was weaker within the central sub-population where kinship was lower overall (Figure S10, Table S10), and we lacked sufficient sites to carry out SAA for northern or southern samples separately. There was no evidence of sex-biased dispersal, based on kinship permutation tests or SAA (Table S10). For Goudet’s sex-biased dispersal tests, only mean variance in \(\:AIc\) and \(\:{F}_{IS}\) were significantly higher among females (p < 0.05). Habitat suitability influences gene flow across the landscape Results of the EEMS analysis were similar between different deme sizes of n = 50 to n = 500. Using n = 100 demes (approximately 1 deme per 2.25 km 2 ), the estimated effective migration rate was significantly higher than the average rate of migration within both the Central and South subpopulations, and the migration rate was significantly lower than expected between these two subpopulations (Figs. 2 , S11). All single surface models tested with resistanceGA outperformed the geographic distance only and null models ( \(\:{\varDelta\:AIC}_{c}\) > 2, Tables 1 , S11). The best supported single surface model was habitat suitability ( \(\:{AIC}_{c}\) = -33,801.07), with areas of suitable habitat having the lowest resistance to gene flow, i.e., Eucalypt dominated vegetation on alluvial soils, Eucalypt woodland on paleo-alluvial sediments and Eucalypt/ Acacia vegetation on metamorphosed sediments (Fig. 2 ). The habitat suitability model was the best supported single surface model 99.3% of the time (out of 1,000 bootstrap iterations). The next best single surface models were canopy cover ( \(\:{\varDelta\:AIC}_{c}\) = 13.7) and wet season distance to water ( \(\:{\varDelta\:AIC}_{c}\) = 99.6) (Table 1 ). For canopy cover, open woodland (6–11% plant cover fraction) and woodland (11–30%) had the lowest resistance to gene flow, while scattered trees ( 30%) had higher resistance. Wet season distance to water had better support than the dry season model, and a similar pattern was observed for the bare earth models (Table 1 ), indicating that the availability of water and vegetation during the wet season had a stronger influence on gene flow relative to the dry season. Table 1 Best run results of the single resistance surface optimization procedure implemented by ResistanceGA for black-throated finch subpopulations around the Carmichael Coal Mine. Single surface models are listed in order of increasing \(\:{AIC}_{c}\) . \(\:{AIC}_{c}\) is the average \(\:{AIC}_{c}\) value obtained for each model in 1,000 bootstrap iterations; \(\:k\) is the number of parameters; \(\:{\varDelta\:AIC}_{c}\) is the difference in \(\:{AIC}_{c}\) between the top model (lowest \(\:{AIC}_{c}\) ) and each subsequent model; and \(\:{mR}^{2}\) is the average marginal \(\:{R}^{2}\) of 1,000 bootstrap iterations. Predictor Best Run \(\:\varvec{k}\) \(\:{\varvec{A}\varvec{I}\varvec{C}}_{\varvec{c}}\) \(\:{\varvec{\varDelta\:}\varvec{A}\varvec{I}\varvec{C}}_{\varvec{c}}\) \(\:{\varvec{m}\varvec{R}}^{2}\) Habitat suitability 1 3 -33801.07 0 0.170 Canopy cover 2 7 -33787.37 13.70 0.167 Wet season distance to water 2 4 -33701.52 99.55 0.233 Wet season bare earth 1 4 -33700.37 100.71 0.242 Dry season bare earth 1 4 -33674.96 126.11 0.217 Canopy height 2 6 -33648.01 153.06 0.129 Dry season distance to water 2 4 -33638.93 162.14 0.159 Geographic distance NA 2 -33613.63 187.45 0.100 Null model NA 1 -33035.88 765.19 0.000 The multi-surface resistance model with the strongest support was a combination of habitat suitability and wet season distance to water ( \(\:{AIC}_{c}\) = -18,649.5, \(\:{W}_{AIC}\) = 1.00). However, this model was only the top model 28.2% of the time out of 1,000 pseudo-bootstrap iterations (Table S12), with the single surface model habitat suitability ranking as the top model 53.1% of the time. In all multi-surface models tested, habitat suitability was identified as the most important variable, contributing > 50% to each combined model. The next best multi-surface model was habitat suitability*wet season distance to water*wet season bare earth ( \(\:{\varDelta\:AIC}_{c}\) = 4.18) (Table S12). Relatively small contemporary \(\:{\varvec{N}}_{\varvec{e}}\) Mean estimates of \(\:{N}_{e}\) across several recent generations ranged from 1,150 to 1,440 based on a \(\:{P}_{crit}\) of 0.02 and depending on the life history traits used for bias corrections (lower 95% C.I. ranged from 940 to 1,180, upper 95% C.I. 1,470 to 1,840) (Table S13). When considering South and North-Central groups as two distinct genetic units, \(\:{N}_{e}\) for the South group (n = 20) ranged from 290 to 360 (lower 95% C.I. from 210 to 260; upper 95% C.I. from 490 to 610) for \(\:{P}_{crit}\) = 0.03, based on the \(\:{P}_{crit}\) rule of thumb proposed by Waples and Do ( 2010 ) for n ≤ 25 ( \(\:1/2n<{P}_{crit}<1/n\) ). For the North-Central group (n = 87), \(\:{N}_{e}\) ranged from 1,280 to 1,610 (lower 95% C.I. 1,010 to 1,270, upper 95% C.I. 1,750 to 2,200; \(\:{P}_{crit}\) = 0.02). Using n = 20, \(\:{N}_{e}\) for the North-Central group was at least twice as large as the South in 90% of bootstrap replicates. Across all life-history traits tested, the average bootstrapped \(\:{N}_{e}\) for the North-Central subpopulation ranged from 970 to 1,220. Evidence of two historical genetic bottlenecks We carried out demographic history analysis for all samples pooled together (‘unrelated’ dataset, subset of 20 samples per region). Although we found evidence of fine-scale genetic structure, we assumed the population was panmictic overall with respect to allele frequencies, given the relatively low global \(\:{F}_{ST}\) and pairwise \(\:{F}_{ST}\) between regions (< 0.02), suggesting drift connectivity (Lowe & Allendorf, 2010 ). The Stairway plot analysis inferred a historical population expansion beginning ≈ 367,000 years ago and lasting ≈ 127,000 years (2 \(\:\times\:\) increase from ancestral \(\:{N}_{e}\) ) (Fig. 3 ). \(\:{N}_{e}\) stabilized until a bottleneck beginning ≈ 700 years ago (82% decrease) followed by a more recent bottleneck ≈ 100 years ago (further 63% decrease from the first bottleneck) (Fig. 3 ). With a longer generation length of 3.5 years, the expansion period detected by Stairway began ≈ 524,000 years ago, the 1st bottleneck ≈ 1,600 years ago and the 2nd bottleneck ≈ 200 years ago. The best fit model in δaδi was the 3 epoch model (3EM, \(\:\varDelta\:AIC\) > 10, Table S14) which inferred a historical population expansion beginning ≈ 321,000 years ago (95% C.I. = 103,000–539,000) to around 2 \(\:\times\:\) the ancestral \(\:{N}_{e}\) , and a recent bottleneck ≈ 60 years ago (95% C.I. = 50–70) (Fig. 3 , Table S15). δaδi detected the most recent population decline compared to Stairway, with \(\:{N}_{e}\) dropping to ≈ 900 within the last 30 generations, similarly low to the estimate by NeEstimator (1,190), which reflects the \(\:{N}_{e}\) of the past few generations. With \(\:g\) = 3.5 years, the expansion began ≈ 562,000 years ago and the bottleneck ≈ 100 years ago. Discussion The endemic and endangered southern black-throated finch has experienced severe range contractions and habitat fragmentation due to anthropogenic processes such as agriculture and land clearing (Reside et al., 2019 ). In this study, we assessed genetic connectivity in one of the last remaining stronghold populations in the Desert Uplands Bioregion. We found genetic structuring over relatively short distances and limited effective dispersal that was predominantly influenced by the availability of suitable habitat. We also detected the genetic signatures of two historical bottlenecks and relatively small contemporary effective population sizes ( \(\:{N}_{e}\) near or below 1000), suggesting BTFS are at risk of losing genetic diversity and adaptive potential in the future. BTFS exhibited restricted recent gene flow over relatively short distances (16–45 km) and lower effective migration between the North-Central and South regions. Assignment tests indicated mixed ancestry within the South region, unidirectional gene flow from Central to South, and no recent gene flow between North and South regions. Additionally, we found evidence of IBD and regional philopatry, and no first- or second-degree relatives were recorded between finches in different regions. These findings are consistent with previous studies on black-throated finches, as discussed below. BTF are philopatric, known to return to the same nest over multiple years, and juveniles remain within family groups for several months after fledging which may promote site fidelity through social learning (Higgins, 2006 ; Isles, 2007 ; NRA, 2005 ; Zann, 1976 ). Radio tracking data suggest BTFS in the Desert Uplands population are generally sedentary over short timescales, dispersing up to 4.5 km and maintaining stationary home ranges over a 1 month period (van Osta et al., 2024 ). In the Townsville population, restricted gene flow has similarly been observed over relatively short distances (10–20 km) (Hooper et al., 2025 ; Tang, 2016 ; Tang et al., 2017 ). While data on long-distance dispersal in BTFS is limited, dietary diversification may be favoured over long-distance dispersal in response to changes in resource availability (Rechetelo et al., 2016 ; van Osta et al., 2024 ). Limited effective dispersal may have implications for the long-term persistence of the Desert Uplands stronghold population, and for the broader population of BTFS across Queensland. Vegetation communities and water availability likely play a larger role in restricting gene flow than geographic distance. Habitat suitability was the strongest predictor of genetic connectivity across the study area. The habitat separating the North-Central and South regions was characterized by highly fragmented suitable habitat and very low ( 30%). Studies on the Townsville population indicate that the presence of a large waterbody (the Ross River Dam), urbanization, vegetation structure and elevation are key barriers to gene flow among BTFS subpopulations (Hooper et al., 2025 ; Tang, 2016 ). Compared to the Townsville population, subpopulations of Desert Uplands finches are not separated by large or permanent waterbodies, and thus vegetation structure and habitat suitability may be more influential in restricting gene flow in this population. In other granivorous birds, landscape features such as reduced forest cover, increased agricultural land or water barriers are linked to restricted gene flow, including the golden-cheeked warbler ( Dendroica chrysoparia ) (Lindsay et al., 2008 ) and song sparrow ( Melospiza melodia ) (Wilson et al., 2011 ). In the black-capped chickadee ( Poecile atricapillus ), a woodland bird species with a high dispersal propensity, spatial patterns of genetic variation were influenced by small breaks in riparian woodlands that function as dispersal corridors across unsuitable grassland habitat (Adams & Burg, 2015 ). Similarly, for the Desert Uplands BTFS population, the fragmentation of suitable woodland habitat between the North-Central and South regions appears to impede effective dispersal across larger areas of unsuitable grassland habitat. Across the range of the southern black-throated finch, there may be subtle differences between populations in dispersal ecology and landscape barriers that restrict gene flow over relatively short distances. Estimates of contemporary \(\:{N}_{e}\) over a few recent generations were low in the Desert Uplands population of BTFS. \(\:{N}_{e}\) informs the risk of losing genetic diversity due to drift over generations (Luikart et al., 2010 ; Tallmon et al., 2010 ). The \(\:{N}_{e}\) estimates for BTFS are near or below the recommended minimum level of 1,000 needed to maintain evolutionary potential in the long term (Frankham et al., 2014 ), particularly in the southern region ( \(\:{N}_{e}\) ~ 300). Adult census sizes ( \(\:{N}_{c}\) , total number of potential breeders) are poorly known for BTFS. The most recent estimate of \(\:{N}_{c}\) is around 450–1,000 in the largest population (Desert Uplands Bioregion), and 633–1,400 across their range, but these estimates are classified as having ‘low’ reliability (Buosi et al., 2021 ). While there isn’t a simple linear relationship between \(\:{N}_{e}\) and \(\:{N}_{c}\) , \(\:{N}_{e}\) is often smaller than \(\:{N}_{c}\) , with \(\:{N}_{e}/{N}_{c}\) ranging from 0.28 to 0.40 in other finch species (Frankham, 1995 ). Overall, this study provides a baseline for continued monitoring of the genetic health of the Desert Uplands population. The recent demographic history reconstructed by Stairway and δaδi is consistent with known range contractions for BTFS. The extent of occurrence of BTFS has contracted by 50–80% since the late 1900’s, and likely by over 90% since pre-European times (BTFRT, 2007 , 2020 ; Franklin, 1999 ; Reside et al., 2019 ). This coincides with the recent genetic bottleneck detected by Stairway and δaδi (60–100 years ago), and the more distant bottleneck detected by Stairway around 700 years ago. Severe declines in \(\:{N}_{e}\) are common in many species around the start of the last glacial period (LGP, 110,000 ya), including in the medium ground finch ( Geospiza fortis ) (Nadachowska-Brzyska et al., 2015 ). Using WGS data, Hooper et al. ( 2025 ) detected a long-term decline in BTFS beginning around 11,300 years ago, predating European arrival. The Townsville and Desert Uplands (Galilee Basin) populations experienced asymmetric migration until around 2,000 years ago, when migration is thought to have dropped to negligible levels (Hooper et al., 2025 ). In this study, we inferred a population expansion occurring ~ 300,000 to 500,000 years ago, consistent with expansions occurring in other bird species during the early to mid-Pleistocene period (Nadachowska-Brzyska et al., 2015 ). The demographic history of the northern subspecies (BTFN) is characterized by population expansions; and estimates of divergence times between BTF and long-tailed finches (LTF, P. acuticauda and hecki ) range from 600,000–1,000,000 years ago (Hooper et al., 2025 ; Jennings & Edwards, 2005 ). Thus, the expansion event we detected in the Desert Uplands population of BTFS may reflect an earlier signature of population expansion, prior to the genetic isolation of BTFS and BTFN. Collectively, evidence of recent genetic bottlenecks within the Desert Uplands Bioregion and negligible gene flow between the two stronghold populations, suggests the long-term viability of BTFS may be compromised. The black-throated finch (northern and southern subspecies) was downlisted to ‘Least Concern’ on the IUCN Red List in 2016, as the rate of decline did not adhere to Red List criteria (BirdLife International, 2024). Buosi et al. ( 2021 ) proposed that BTFS may be eligible for a higher classification based on category A of the Red List criteria (population decline projected to be > 50–80% in three generations based on declines in habitat quality) and category D (250–1,000 mature individuals and a small, restricted distribution). In our genetic study on BTFS in the largest remaining stronghold population, we found evidence of multiple historical genetic bottlenecks, restricted gene flow over relatively short distances likely driven by the fragmentation of suitable habitat, and small contemporary effective population sizes (290–360 in the South and 1,280 to 1,610 in the North-Central region). Recent evidence obtained with WGS data also suggests the Townsville and Desert Uplands stronghold populations have been genetically isolated for around 2,000 years (Hooper et al., 2025 ). With further rapid declines predicted in the near future, BTFS may warrant a higher and separate sub-species listing on the IUCN Red List (Buosi et al., 2021 ). Currently, RL categories and criteria (version 3.1) do not integrate genetic data (IUCN, 2012 ), despite the improvements that could be made in assessing extinction risks (Garner et al., 2020 ). Incorporating both genetic and ecological data into management plans would improve conservation planning and help prevent continued declines in the last two remaining stronghold populations of the southern black-throated finch. Declarations Acknowledgements The authors would like to thank Bravus Mining and Resources for funding this study. Funding: This study was funded by Bravus Mining and Resources. Competing interests: The authors declare no conflicts of interest. Ethics approval: All methods were carried out under AEC ethics approval CA 2020/07/1392, issued by the Queensland Department of Agriculture and Fisheries; research permit WA0025814, issued by the Queensland Department of Environment, Science and Innovation; and Australian Bird and Bat Banding Scheme authority 2832–01, issued by the Australian Government Department of Climate Change, Energy, the Environment and Water. Consent to participate: Not applicable. Consent for publication: Not applicable. Availability of data and materials: Supporting methods and results are available in the Electronic Supplementary Material. Code availability: Not applicable. Author contributions: SG, SD, JVO, JM and BD conceived the project. SG and AS supervised the project. BD, JVO and JM secured funding and managed data collection. SD conducted all statistical analyses and wrote the manuscript, with minor edits from SG, AS and JVO. WA assisted with landscape resistance analyses. References Adams RV, Burg TM (2015) Gene flow of a forest-dependent bird across a fragmented landscape. PLoS ONE 10(11). https://doi.org/10.1371/journal.pone.0140938. Allendorf FW, Luikart G (2009) Conservation and the genetics of populations. John Wiley & Sons. Baldwin M. (1976). Distribution of the black-throated finch. Australian Birds 11:13-14. Beaumont MA, Balding DJ (2004). Identifying adaptive genetic divergence among populations from genome scans. Mol Ecol 13(4):969-980. Beninde J, Wittische J, Frantz AC (2024) Quantifying uncertainty in inferences of landscape genetic resistance due to choice of individual‐based genetic distance metric. Mol Ecol Resour 24(1). https://doi.org/10.1111/1755-0998.13831. Bergeron LA, Besenbacher S, Zheng J, Li P, Bertelsen MF, Quintard B, Hoffman JI, Li Z, St. Leger J, Shao C (2023) Evolution of the germline mutation rate across vertebrates. Nature 615:285-291. Bezanson J, Edelman A, Karpinski S, Shah VB (2017) Julia: A fresh approach to numerical computing. SIAM Review 59(1):65-98. Bird JP, Martin R, Akçakaya HR, Gilroy J, Burfield IJ, Garnett ST, Symes A, Taylor J, Şekercioğlu ÇH, Butchart SH (2020) Generation lengths of the world's birds and their implications for extinction risk. Conserv Biol 34(5):1252-1261. BirdLife International (2024) Poephila cincta . The IUCN Red List of Threatened Species 2024: e.T22719692A253990937. https://dx.doi.org/10.2305/IUCN.UK.2024-2.RLTS.T22719692A253990937.en. Accessed on 30 May 2025. Brondízio ES, Settele J, Diaz S, Ngo HT (2019) Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES secretariat p.1148. BTFRT (2007) National recovery plan for the black-throated finch southern subpecies (Poephila cincta cincta). Black Throated Finch Recovery Team, Department of Environment and Climate Change (NSW) and Queensland Parks and Wildlife Service. Report prepared for the Department of the Environment and Water Resources, Canberra. Department of Environment and Climate Change (NSW), Hurstville and Queensland Parks and Wildlife Service, Brisbane. https://www.dcceew.gov.au/sites/default/files/documents/p-cincta.pdf. Accessed 30 May 2025. BTFRT (2020) Conversation action plan summary for the black-throated finch southern subspecies. Black Throated Finch Recovery Team, Department of Environment and Climate Change (NSW) and Queensland Parks and Wildlife Service. https://blackthroatedfinch.org/wp-content/uploads/2024/01/BTF-Conservation-Action-Plan-Summary.pdf. Accessed 30 May 2025. Buosi P, Vanderduys EP, Reside AE (2021) Southern black-throated finch Poephila cincta cincta. In Garnett ST and Baker GB (eds), The Action Plan for Australian Birds 2020 (pp. 786-789). CSIRO Publishing. Conomos MP, Thornton T, Gogarten S (2016) GENetic EStimation and inference in structured samples (GENESIS): statistical methods for analyzing genetic data from samples with population structure and/or relatedness. R package version, 2(0.1). Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol Ecol 14(8):2611-2620. Excoffier L, Lischer HE (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour 10(3):564-567. Exposito-Alonso M, Booker TR, Czech L, Gillespie L, Hateley S, Kyriazis CC, Lang PL, Leventhal L, Nogues-Bravo D, Pagowski V (2022) Genetic diversity loss in the Anthropocene. Science 377(6613):1431-1435. Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164(4):1567-1587. Foll M, Gaggiotti OE (2008) A genome scan method to identify selected loci appropriate for both dominant and codominant markers: a Bayesian perspective. Genetics 180(2):977–993. Forshaw JM, Shephard M (2012) Grassfinches in Australia. CSIRO Publishing. Frankham R (1995) Effective population size/adult population size ratios in wildlife: a review. Genet Res 66(2):95-107. Frankham R (2005) Genetics and extinction. Biological Conservation, 126(2), 131-140. Frankham R, Ballou J, Ralls K, Eldridge M, Dudash M, Fenster C, Lacy R, Sunnucks P (2017) Population fragmentation causes inadequate gene flow and increases extinction risk. Genetic management of fragmented animal and plant populations, 87-112. Frankham R, Bradshaw CJ, Brook BW (2014) Genetics in conservation management: revised recommendations for the 50/500 rules, Red List criteria and population viability analyses. Biol Conserv 170:56-63. Franklin DC (1999) Evidence of disarray amongst granivorous bird assemblages in the savannas of northern Australia, a region of sparse human settlement. Biol Conserv 90(1):53-68. Garner BA, Hoban S, Luikart G (2020) IUCN Red List and the value of integrating genetics. Conserv Gen 21(5):795-801. https://doi.org/10.1007/s10592-020-01301-6. Garnett ST, Szabo JK, Duston G (2011) The Action Plan for Australian Birds 2010. CSIRO. Goudet J (2005) Hierfstat, a package for R to compute and test hierarchical F‐statistics. Mol Ecol Notes 5(1):184-186. Goudet J, Perrin N, Waser P (2002) Tests for sex‐biased dispersal using bi‐parentally inherited genetic markers. Mol Ecol 11(6):1103-1114. Grice AC, Large T, Kahler CP, Heading R (2023) Surveys of a small population of the Endangered Black-throated Finch 'Poephila cincta cincta' in North Queensland. Aust Field Ornith 40:152-164. Gruber B, Unmack PJ, Berry OF, Georges A (2018) dartr: An r package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Mol Ecol Resour 18(3):691-699. Gutenkunst RN, Hernandez RD, Williamson SH, Bustamante CD (2009) Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data. PLoS genetics 5(10). https://doi.org/10.1371/journal.pgen.1000695. Hall KR, Anantharaman R, Landau VA, Clark M, Dickson BG, Jones A, Platt J, Edelman A, Shah VB (2021) Circuitscape in Julia: empowering dynamic approaches to connectivity assessment. Land 10(3). https://doi.org/10.3390/land10030301. Higgins P (2006) Handbook of Australian, New Zealand and Antarctic Birds: Volume 7 Boatbill to Starlings: Part B Dunnock to Starlings. Oxford University Press. Hoban S, Archer FI, Bertola LD, Bragg JG, Breed MF, Bruford MW, Coleman MA, Ekblom R, Funk WC, Grueber CE (2022) Global genetic diversity status and trends: towards a suite of Essential Biodiversity Variables (EBVs) for genetic composition. Biol Rev 97(4):1511-1538. Hoban S, Bruford MW, Funk WC, Galbusera P, Griffith MP, Grueber CE, Heuertz M, Hunter ME, Hvilsom C, Stroil BK (2021) Global commitments to conserving and monitoring genetic diversity are now necessary and feasible. BioScience 71(9):964-976. Hoban S, Campbell CD, da Silva JM, Ekblom R, Funk WC, Garner BA, Godoy JA, Kershaw F, MacDonald AJ, Mergeay J (2021) Genetic diversity is considered important but interpreted narrowly in country reports to the Convention on Biological Diversity: Current actions and indicators are insufficient. Biol Conserv 261. https://doi.org/10.1016/j.biocon.2021.109233. Holderegger R, Balkenhol N, Bolliger J, Engler JO, Gugerli F, Hochkirch A, Nowak C, Segelbacher G, Widmer A, Zachos FE (2019) Conservation genetics: linking science with practice. Mol Ecol 28(17):3848-3856. Hooper DM, Lopez KA, Butcher BG, Lovette IJ, Griffith SC (2025) Recent divergence and microgeographic genetic structure in an endangered Australian songbird: the southern black-throated finch. bioRxiv. https://doi.org/10.1101/2025.04.17.649380. Huang X, Struck TJ, Davey SW, Gutenkunst RN (2023) dadi-cli: Automated and distributed population genetic model inference from allele frequency spectra. bioRxiv. https://doi.org/10.1101/2023.06.15.545182. Iacchei M, Ben‐Horin T, Selkoe KA, Bird CE, García‐Rodríguez FJ, Toonen RJ (2013) Combined analyses of kinship and FST suggest potential drivers of chaotic genetic patchiness in high gene‐flow populations. Mol Ecol, 22(13):3476-3494. Isles NL (2007) Patterns of Foraging Behaviour of the Blackthroated Finch during Breeding Season. Dissertation, James Cook University, Townsville, Queensland. IUCN (2012) IUCN Red List Categories and Criteria: Version 3.1. Second edition. International Union for Conservation of Nature and Natural Resources. Jakobsson M, Rosenberg NA (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinf 23(14):1801-1806. Jennings WB, Edwards SV (2005) Speciational history of Australian grass finches (Poephila) inferred from thirty gene trees. Evolution 59(9):2033-2047. Jombart T (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinf 24(11):1403-1405. Jombart T, Ahmed I (2011) adegenet 1.3-1: new tools for the analysis of genome-wide SNP data. Bioinf 27(21):3070-3071. Keenan K, McGinnity P, Cross TF, Crozier WW, Prodöhl PA (2013) diveRsity: an R package for the estimation and exploration of population genetics parameters and their associated errors. Methods Ecol Evol 4(8):782-788. Kilian A, Wenzl P, Huttner E, Carling J, Xia L, Blois H, Caig V, Heller-Uszynska K, Jaccoud D, Hopper C (2012) Diversity arrays technology: a generic genome profiling technology on open platforms. In Pompanon F & Bonin A (eds) Data Production and Analysis in Population Genomics. Humana Press, pp 67-89. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, Subgroup GPDP (2009) The sequence alignment/map format and SAMtools. Bioinf 25(16):2078-2079. Lindsay DL, Barr KR, Lance RF, Tweddale SA, Hayden TJ, Leberg PL (2008) Habitat fragmentation and genetic diversity of an endangered, migratory songbird, the golden-cheeked warbler (Dendroica chrysoparia). Mol Ecol 17(9):2122-2133. https://doi.org/10.1111/j.1365-294X.2008.03673.x Liu S, Zeng Y, Wang C, Zhang Q, Chen M, Wang X, Wang L, Lu Y, Guo H, Bu F (2022) seGMM: A New Tool for Gender Determination From Massively Parallel Sequencing Data. Front Genet 13. https://doi.org/10.3389/fgene.2022.850804. Liu X, Fu YX (2015) Exploring population size changes using SNP frequency spectra. Nat Genet 47(5):555-559. Lowe WH, Allendorf FW (2010) What can genetics tell us about population connectivity? Mol Ecol 19(15):3038-3051. Luikart G, Ryman N, Tallmon DA, Schwartz MK, Allendorf FW (2010) Estimation of census and effective population sizes: the increasing usefulness of DNA-based approaches. Conserv Genet 11:355-373. Miller JM, Cullingham CI, Peery RM (2020) The influence of a priori grouping on inference of genetic clusters: simulation study and literature review of the DAPC method. Heredity 125(5):269-280. Mula Laguna J, Reside AE, Kutt A, Grice AC, Buosi P, Vanderduys EP, Taylor M, Schwarzkopf L (2019) Conserving the endangered Black-throated Finch southern subspecies: what do we need to know? Emu - Austral Ornithol 119(4):331-345. Nadachowska-Brzyska K, Li C, Smeds L, Zhang G, Ellegren H (2015) Temporal dynamics of avian populations during Pleistocene revealed by whole-genome sequences. Curr Biol 25(10):1375-1380. NRA (2005) 'Enertrade North Queensland Gas Pipeline - Black-throated finch studies (post-construction)' Report prepared by NRA Environmental Consultants to Enertrade. Peterman WE (2018) ResistanceGA: An R package for the optimization of resistance surfaces using genetic algorithms. Methods Ecol Evol 9(6):1638-1647. Petkova D (2024) reemsplots2: Generate plots to inspect and visualize the results of EEMS. R package version 0.1.0, commit fbc2de644d9f591c91b795489b5026f9851f7fd8, . Petkova D, Novembre J, Stephens M (2016) Visualizing spatial population structure with estimated effective migration surfaces. Nat Genet 48(1):94-100. Pierson JC, Coates DJ, Oostermeijer JGB, Beissinger SR, Bragg JG, Sunnucks P, Schumaker NH, Young AG (2016) Genetic factors in threatened species recovery plans on three continents. Front Ecol Environ 14(8):433-440. Qian DC, Busam JA, Xiao X, O’Mara TA, Eeles RA, Schumacher FR, Phelan CM, Amos CI (2017) seXY: a tool for sex inference from genotype arrays. Bioinf 33(4):561-563. Raffard A, Santoul F, Cucherousset J, Blanchet S (2019) The community and ecosystem consequences of intraspecific diversity: A meta‐analysis. Biol Rev 94(2):648-661. Rechetelo J (2015) Movement, habitat requirements, nesting and foraging site selection: a case study of an endangered granivorous bird, the Black-throated finch Poephila cincta cincta in north-eastern Australia. Dissertation, James Cook University. Rechetelo J, Grice A, Reside AE, Hardesty BD, Moloney JM (2016) Movement patterns, home range size and habitat selection of an endangered resource tracking species, the black-throated finch (Poephila cincta cincta). PLoS ONE 11(11). doi:0167210.0161371/journal.pone.0167254. Reside AE, Beher J, Cosgrove AJ, Evans MC, Seabrook L, Silcock JL, Wenger AS, Maron M (2017) Ecological consequences of land clearing and policy reform in Queensland. Pacific Conserv Biol 23(3):219-230. Reside AE, Cosgrove AJ, Pointon R, Trezise J, Watson JE, Maron M (2019) How to send a finch extinct. Environmental Science & Policy 94:163-173. Rougeux C, Bernatchez L, Gagnaire PA (2017) Modeling the multiple facets of speciation-with-gene-flow toward inferring the divergence history of lake whitefish species pairs (Coregonus clupeaformis). Genome Biol Evol 9(8):2057-2074. Sansaloni C, Petroli C, Jaccoud D, Carling J, Detering F, Grattapaglia D, Kilian A (2011) Diversity Arrays Technology (DArT) and next-generation sequencing combined: genome-wide, high throughput, highly informative genotyping for molecular breeding of Eucalyptus. In BMC proceedings (Vol. 5, No. Suppl 7, p. P54). London: BioMed Central. Schodde R, Mason IJ (1999) Directory of Australian birds: passerines: Passerines. CSIRO publishing. Schubert M, Lindgreen S, Orlando L (2016) AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res Notes 9:1-7. Scrucca L, Fop M, Murphy TB, Raftery AE (2016) mclust 5: clustering, classification and density estimation using Gaussian finite mixture models. The R Journal 8(1): p 289. Shephard M (1989) Aviculture in Australia: Keeping and Breeding Aviary Birds. Black Cockatoo Press. Shirk A, Landguth E, Cushman S (2017) A comparison of individual‐based genetic distance metrics for landscape genetics. Mol Ecol Resour 17(6):1308-1317. Smouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82(5):561-573. Tallmon DA, Gregovich D, Waples RS, Baker CS, Jackson J, Taylor BL, Archer E, Martien KK, Allendorf FW, Schwartz MK (2010) When are genetic methods useful for estimating contemporary abundance and detecting population trends? Mol Ecol Resour 10(4):684-692. Tang LS (2016) Conservation genetics of granivorous birds in a heterogeneous landscape: the case of the black-throated finch (Poephila cincta). Dissertation, James Cook University. Tang LS, Smith-Keune C, Grice AC, Moloney JM, Hardesty BD (2017) Genetic structure and diversity of the black-throated finch (Poephila cincta) across its current range. Aust J Zool 64(6):375-384. van Osta JM, Dreis B, Grogan LF, Castley JG (2024) Local resource availability drives habitat use by a threatened avian granivore in savanna woodlands. PLoS ONE 19(8). https://doi.org/10.1371/journal.pone.0306842. Vanderduys EP, Reside AE, Grice A, Rechetelo J (2016) Addressing potential cumulative impacts of development on threatened species: the case of the endangered black-throated finch. PLoS ONE 11(3). https://doi.org/10.1371/journal.pone.0148485. Walters AD, Schwartz MK (2020) Population genomics for the management of wild vertebrate populations. In: Population Genomics: Wildlife. Cham: Springer International Publishing, pp 419-436. Waples RS, Do C (2008) LDNE: a program for estimating effective population size from data on linkage disequilibrium. Mol Ecol Resour 8(4):753-756. Waples RS, Do C (2010) Linkage disequilibrium estimates of contemporary Ne using highly variable genetic markers: a largely untapped resource for applied conservation and evolution. Evol Appl 3(3):244-262. Whitlock MC, Lotterhos KE (2015) Reliable detection of loci responsible for local adaptation: inference of a null model through trimming the distribution of F ST. Am Nat 186(1):24-36. Wigginton JE, Cutler DJ, Abecasis GR (2005) A note on exact tests of Hardy-Weinberg equilibrium. Am J Hum Genet 76(5):887-893. Wilson AG, Arcese P, Chan YL, Patten MA (2011) Micro-spatial genetic structure in song sparrows (Melospiza melodia). Conserv Genet 12:213-222. Woinarski J, Catterall C (2004) Historical changes in the bird fauna at Coomooboolaroo, northeastern Australia, from the early years of pastoral settlement (1873) to 1999. Biol Conserv 116(3):379-401. Zann R (1976) Distribution, status and breeding of Black-throated Finches Poephila cincta in northern Queensland. Emu – Austral Ornithol 76(4):201-206. Zheng X, Zheng MX (2013) Package ‘SNPRelate’. A package for parallel computing toolset for relatedness and principal component analysis of SNP data. Additional Declarations No competing interests reported. Supplementary Files DavisetalESM.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 23 Sep, 2025 Reviewers agreed at journal 10 Jun, 2025 Reviewers invited by journal 08 Jun, 2025 Editor assigned by journal 02 Jun, 2025 Submission checks completed at journal 02 Jun, 2025 First submitted to journal 30 May, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6780931","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":468303061,"identity":"2536b357-a32a-4415-81c9-4b6a9ff51956","order_by":0,"name":"Skye Davis","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYBACCQYGZhDN2ABCH4AsPqgIcVoYZwAF2EjQAmTxEKNFsoH9sTFPBYNs/4zk5s82FYfr2NibDxsw1NhE49IizcBjnMxzhsF4xo3ENumcM4cl2HiOJScwHEvLbcChRY6Bh/kwbxtDYsOZg23MuW1ALRI5xgcYGw7j0cL++DDvP4bE+WcONn+2JEaLNAODcTJvA0PihuONDdKMUC0J+LRINvMYG845JmG88Xhjm2TPmXTJNqBfDBLw+EXiePtjiTc1NrLzDrM//vCjwpqfHxhiEh9qbHBqAcUBEw8oelBAAi7lUMD4g4CCUTAKRsEoGOEAAHo7Uc3L+VXfAAAAAElFTkSuQmCC","orcid":"","institution":"Macquarie University","correspondingAuthor":true,"prefix":"","firstName":"Skye","middleName":"","lastName":"Davis","suffix":""},{"id":468303062,"identity":"fd372cad-b987-43f5-a364-ff1b6e19dbe9","order_by":1,"name":"Adam J Stow","email":"","orcid":"","institution":"Macquarie University","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"J","lastName":"Stow","suffix":""},{"id":468303063,"identity":"855e6a8c-eea9-490b-aca8-0bd34ec48fde","order_by":2,"name":"Jemma McCrossin","email":"","orcid":"","institution":"E2M Pty Ltd","correspondingAuthor":false,"prefix":"","firstName":"Jemma","middleName":"","lastName":"McCrossin","suffix":""},{"id":468303064,"identity":"a5775a69-4a97-41d9-ad41-d0587b9055c9","order_by":3,"name":"Wilbur Ashley","email":"","orcid":"","institution":"Macquarie University","correspondingAuthor":false,"prefix":"","firstName":"Wilbur","middleName":"","lastName":"Ashley","suffix":""},{"id":468303066,"identity":"3d483c9d-891a-4b24-be48-70459468bea1","order_by":4,"name":"John M van Osta","email":"","orcid":"","institution":"E2M Pty Ltd","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"M van","lastName":"Osta","suffix":""},{"id":468303067,"identity":"ce92c2da-7696-4794-90fc-aaf28fb1fd4f","order_by":5,"name":"Brad Dreis","email":"","orcid":"","institution":"E2M Pty Ltd","correspondingAuthor":false,"prefix":"","firstName":"Brad","middleName":"","lastName":"Dreis","suffix":""},{"id":468303068,"identity":"9b04190c-d564-430b-a30b-0d4ea306abfc","order_by":6,"name":"Simon Griffith","email":"","orcid":"","institution":"Macquarie University","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Griffith","suffix":""}],"badges":[],"createdAt":"2025-05-30 04:39:05","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6780931/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6780931/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":84289090,"identity":"27e7d9d8-eee3-4540-bc97-c1be7d191b87","added_by":"auto","created_at":"2025-06-10 08:16:45","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":147119,"visible":true,"origin":"","legend":"\u003cp\u003eA) Map of southern black-throated finch samples collected around the Carmichael Coal Mine and coloured by region (blue = North, red = Central, green = South). B) Discriminant analysis of principal components (DAPC) showing the first two discriminant functions, with individuals coloured by region of origin. C) Spatial autocorrelation analysis (SAA) showing mean within distance-class genotypic correlation (related to kinship). Irregular distance classes were chosen to maximise sample size per distance class, solid error bars denote 95% bootstrapped C.I. about the mean r within each distance class (999 bootstraps), dashed line denotes the 95% C.I. for the hypothesis of ‘no spatial structure’ (999 permutations). D) STRUCTURE plot of admixture coefficients for each individual, assuming 𝐾 = 3 ancestral populations (highest delta 𝐾).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6780931/v1/b31c3de48b704372fee0fb27.png"},{"id":84290099,"identity":"922560ae-4bcd-48df-b5cd-35bd24e473bc","added_by":"auto","created_at":"2025-06-10 08:24:45","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":571579,"visible":true,"origin":"","legend":"\u003cp\u003eResults of the functional connectivity analyses for black-throated finches located around the Carmichael Coal Mine. A) Sample locations coloured by region (blue = North, red = Central, green = South), with the number in each circle indicating the number of samples in that area. B-D) Landscape layers for the top three single-surface resistance models in ResistanceGA, prior to optimization, including B) habitat suitability based on vegetation classes for BTF, C) percentage tree canopy cover grouped into 6 classes and D) minimum distance (km) to water during the wet season, including bores, dams, waterholes, ephemeral and permanent natural water. E) Effective migration rate (log(m)) across the study area as estimated with EEMS, showing areas of higher (blue) or lower (orange) than expected migration rates. F-H) Least-cost resistance pathways generated with Circuitscape, showing current flow between samples across each optimized resistance surface.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6780931/v1/cf00104b0410822486096ad4.png"},{"id":84289086,"identity":"9e7c47df-3d56-4110-aba3-e95f3e5a3503","added_by":"auto","created_at":"2025-06-10 08:16:45","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":89975,"visible":true,"origin":"","legend":"\u003cp\u003eStairway plot showing historical changes in the median effective population size (𝑁𝑒) of black-throated finches in the southern Desert Uplands Bioregion, over A) the last million years and B) the last 1,000 years, with shaded polygons representing the 95% C.I.’s (light grey) and 75% C.I.’s (dark grey). Demographic events inferred by δaδi (best-fit model = 3 epoch) are indicated in orange. Results for both approaches are shown based on a mutation rate per site per generation of 5.85×10-9 and a generation length (𝑔) of 2 years. For results with 𝑔 = 3.5 years, see Figure S12.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6780931/v1/7efee11dea75bee552be1e4c.png"},{"id":84291216,"identity":"e5fc47dc-e163-466f-a356-0c55daf936f9","added_by":"auto","created_at":"2025-06-10 08:40:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1658805,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6780931/v1/89e53df9-9939-40aa-9f42-8ebe9fb0fc53.pdf"},{"id":84290105,"identity":"0fa05501-2bfa-42d0-99a1-a99a089a4e8b","added_by":"auto","created_at":"2025-06-10 08:24:46","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":14004662,"visible":true,"origin":"","legend":"","description":"","filename":"DavisetalESM.docx","url":"https://assets-eu.researchsquare.com/files/rs-6780931/v1/599ad9a02a2fb1763ed93001.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Recent historical bottlenecks and restricted gene flow in one of the last remaining stronghold populations of the southern black-throated finch (Poephila cincta cincta)","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOngoing species declines are occurring despite global conservation efforts and environmental protection laws. Biodiversity is being lost at an unprecedented scale, and the current extinction rate is higher than the background rate for many taxonomic groups, particularly birds, amphibians and mammals (Brond\u0026iacute;zio et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). As the fundamental source of biodiversity, conserving genetic diversity in wild populations is vital for supporting the long-term persistence of species (Allendorf \u0026amp; Luikart, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Lowe \u0026amp; Allendorf, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Raffard et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Reduced gene flow, for example due to habitat fragmentation, can lead to a loss of genetic diversity and an increased risk of inbreeding depression, reducing adaptive potential and elevating extinction risk (Frankham, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Frankham et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). More than 10% of global genetic diversity may have already been lost based on the relationship between decreasing habitat and the loss of genome-wide diversity (Exposito-Alonso et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and a 19\u0026ndash;66% decline in genetic diversity is forecasted without intervention (Hoban, Bruford, et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Consequently, molecular approaches have been widely adopted in wildlife management, particularly with the rise of high-throughput genotyping technologies and whole genome sequencing (Hoban et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Holderegger et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Pierson et al., \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Despite these advances, there is a severe lag in incorporating genomic data into management plans and policies, and challenges remain with how genetic data is utilized in an effective and timely manner for conservation (Hoban, Campbell, et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Walters \u0026amp; Schwartz, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe black-throated finch (\u003cem\u003ePoephila cincta\u003c/em\u003e, hereafter referred to as BTF) is a granivorous (seed-eating) finch endemic to Australia with two recognised subspecies, a northern (\u003cem\u003eP. c. atropygialis\u003c/em\u003e, BTFN) and southern form (\u003cem\u003eP. c. cincta\u003c/em\u003e, BTFS) (Higgins, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). BTF have declined since the early 1900\u0026rsquo;s with the rise of pastoralism, and their range has contracted by more than 80% since the late 1900\u0026rsquo;s (BTFRT, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Franklin, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The southern subspecies was formerly distributed from north-eastern NSW up to the Atherton Tablelands (Baldwin, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1976\u003c/span\u003e), but its range is now highly fragmented and restricted to north-eastern QLD (Schodde \u0026amp; Mason, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). Two strongholds remain; the Townsville Coastal Plain and the Desert Uplands Bioregion (Galilee Basin), where the largest remaining population is found (Grice et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Vanderduys et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Although BTF is listed as \u0026lsquo;Least Concern\u0026rsquo; on the IUCN Red List of Threatened Species (BirdLife International, 2024), BTFS was nationally listed in 2005 as \u0026ldquo;Endangered\u0026rdquo; (Environment Protection and Biodiversity Conservation Act 1999), and state-listed in 2016 as \u0026ldquo;Presumed Extinct\u0026rdquo; in NSW (Threatened Species Conservation Act 1995). Yet despite increased environmental regulation since BTFS was first listed under these environmental protection laws, the southern subspecies has continued to decline in recent years (Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Conserving the last two strongholds is crucial for ensuring the long-term persistence of BTFS.\u003c/p\u003e \u003cp\u003eThe major threats facing BTFS are habitat loss and degradation, particularly in the south of its range where agriculture has been more intensive (Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). BTF have strict habitat requirements, occupying open grassy woodlands and forests with regular access to drinking water (Higgins, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Zann, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e1976\u003c/span\u003e). Based on flock sizes and radio tracking data, BTFS generally prefer habitats close to permanent water sources, with low tree and shrub cover and a higher cover and richness of native grasses (higher vegetation density within the 0\u0026ndash;20 cm height class), avoiding heavily grazed areas with low grass diversity (Rechetelo, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Rechetelo et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The decline of BTFS has been primarily driven by agriculture, clearing of grassy woodlands, habitat degradation by livestock grazing and the spread of invasive species (e.g., exotic grasses) (BTFRT, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Forshaw \u0026amp; Shephard, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Garnett et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Woinarski \u0026amp; Catterall, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). More than half of the remaining habitat for BTFS is threatened by mining, particularly affecting the larger resident population in the Desert Uplands Bioregion (Vanderduys et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The Townsville population is predominantly threatened by urbanization and land development (BTFRT, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Mula Laguna et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Flock sizes have decreased in both populations (Mula Laguna et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), but are larger in the Desert Uplands Bioregion, likely due to a higher percentage of remnant vegetation and intact BTF habitat compared to Townsville (Reside et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). BTFS are also considered sedentary, maintaining small home range sizes and exhibiting limited post-breeding dispersal, which may exacerbate their risk of location extinction (Garnett et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Higgins, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Rechetelo et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eRecovery efforts over the past two decades have aimed at improving ecological knowledge available for BTFS to inform conservation status and guide management strategies (BTFRT, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Yet conservation remains hampered by limited information on its ecology and a lack of long-term monitoring in the face of ongoing habitat loss (Mula Laguna et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Information on the movement ecology of BTFS has largely been based on data from the Townsville population (Mula Laguna et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), although more recently, radio tracking was used to investigate habitat use, home range and daily movements in the Desert Uplands population (van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). There have also been relatively few genetic studies on BTFS. A recent whole-genome sequencing (WGS) study supports the distinction of two subspecies of BTF, suggesting the Einasleigh Uplands restricts gene flow between BTFN and BTFS with no evidence of recent admixture (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). A fine-scale study of the Townsville BTFS population, based on microsatellite markers, found genetic structuring over relatively short distances (10\u0026ndash;20 km), with differences in elevation and vegetation structure, along with the Ross River Dam, contributing to altered gene flow (Tang, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tang et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). WGS data also supports strong spatial genetic structure within the Townsville population, as well as genetic differentiation between the Townsville and Desert Uplands (Galilee Basin) populations (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, further research is needed to examine genetic structure and diversity within the larger, resident Desert Uplands population of southern black-throated finch and assess potential genetic consequences of recent population declines. Genetic data can help improve population viability models for BTFS and assess the long-term persistence of this subspecies in the face of environmental change.\u003c/p\u003e \u003cp\u003eHere, we use a panel of genome-wide single nucleotide polymorphisms (SNPs) to investigate the functional connectivity and genetic health of southern black-throated finches (\u003cem\u003eP. c. cincta\u003c/em\u003e) within the Desert Uplands Bioregion population centred around the Carmichael Coal Mine (CCM). We aim to 1) assess spatial genetic structure, diversity and site fidelity; 2) identify potential biogeographic barriers to dispersal with landscape resistance modelling; and 3) estimate effective population size and infer demographic history.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSample collection\u003c/h2\u003e \u003cp\u003eBlood samples were collected from black-throated finches on the Moray Downs property (Bravus Mining and Resources) in the Desert Uplands Bioregion of Queensland, Australia, between 2021 and 2023 (Williams et al., \u003cem\u003ein prep\u003c/em\u003e). Samples were collected from three localities (denoted North, Central and South regions, see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSequencing, SNP calling and filtering\u003c/h3\u003e\n\u003cp\u003e158 BTFS blood samples were sent to Diversity Arrays Technology (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.diversityarrays.com\u003c/span\u003e\u003cspan address=\"http://www.diversityarrays.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) for DArTseq genotyping-by-sequencing (GBS) on an Illumina platform (Kilian et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Sansaloni et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Library construction was carried out using the DArTseq complexity reduction approach. Genomic DNA was digested with \u003cem\u003ePstI\u003c/em\u003e and \u003cem\u003eSp4\u003c/em\u003e restriction enzymes, and barcoded adapter-ligated fragments were amplified with PCR. Libraries were sequenced on an Illumina HiSeq\u0026reg;2500 platform following manufacturer instructions, and samples were demultiplexed according to their ligated barcode. Raw reads were processed to remove barcodes using \u003cem\u003eFASTX-Toolkit\u003c/em\u003e v 0.11.8 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://hannonlab.cshl.edu/fastx_toolkit/\u003c/span\u003e\u003cspan address=\"http://hannonlab.cshl.edu/fastx_toolkit/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), Illumina adapters were removed using AdapterRemoval v 2.3.1 (Schubert et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), and reads with Phred Q-scores\u0026thinsp;\u0026gt;\u0026thinsp;25 were retained. Quality was checked with \u003cem\u003eFastQC\u003c/em\u003e v 0.11.7 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bioinformatics.babraham.ac.uk/projects/fastqc/\u003c/span\u003e\u003cspan address=\"http://www.bioinformatics.babraham.ac.uk/projects/fastqc/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor analysis of neutral genetic structure, reads were checked for contamination against the Bacteria NCBI database and processed using DArT\u0026rsquo;s proprietary pipeline DArTSoft14\u0026trade; (Diversity Arrays Technology P/L). SNPs were called \u003cem\u003ede novo\u003c/em\u003e by first calling sequencing clusters from pooled samples and then individual samples, removing monomorphic sequence clusters. SNPs were filtered with the R package \u003cem\u003edartR\u003c/em\u003e v 2.7.2 (Gruber et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) by applying the following filters: average reproducibility\u0026thinsp;\u0026gt;\u0026thinsp;0.98; mean read depth \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:0\u0026lt;rd\u0026lt;rd+3\\times\\:S.D.\\left(rd\\right)\\)\u003c/span\u003e\u003c/span\u003e; locus call rate\u0026thinsp;\u0026gt;\u0026thinsp;0.95; sample call rate\u0026thinsp;\u0026gt;\u0026thinsp;0.90; minor allele count (MAC) \u0026ge; 3; retaining one SNP per RAD tag at random; removal of loci that deviated from HWE within populations using the exact tests of Wigginton et al. (\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) with Bonferroni corrections; and removal of duplicate samples with \u003cem\u003eSNPRelate\u003c/em\u003e v 1.34.1 (Zheng \u0026amp; Zheng, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) where kinship coefficient exceeded 0.45. To generate a set of putatively neutral SNPs, any loci detected as outliers by three independent methods were removed, \u003cem\u003eOutFLANK\u003c/em\u003e v 0.2 (Whitlock \u0026amp; Lotterhos, \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), BayeScan v 2.1 (Beaumont \u0026amp; Balding, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Foll \u0026amp; Gaggiotti, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) and Arlequin v 3.5 (Excoffier \u0026amp; Lischer, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Close relatives were identified and removed to obtain an \u0026lsquo;unrelated\u0026rsquo; dataset, using the functions \u0026lsquo;pcrelate\u0026rsquo; and \u0026lsquo;pcairPartition\u0026rsquo; from GENESIS v 2.32.0 to update KING kinship coefficients in the presence of population stratification (Conomos et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The \u0026lsquo;unrelated\u0026rsquo; dataset was used for most downstream analyses of genetic structure, owing to the substantial impact close relatives had on genetic structure (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The \u0026lsquo;related\u0026rsquo; dataset (all samples) was used for identifying recent migrants between populations.\u003c/p\u003e \u003cp\u003eFor analysis of historical demography, raw reads were further processed to retain reads with Q-scores\u0026thinsp;\u0026gt;\u0026thinsp;33 and were mapped to a reference genome using \u003cem\u003eBowtie2\u003c/em\u003e v 2.4.5, \u003cem\u003ePicard\u003c/em\u003e v 2.26.2 and \u003cem\u003eSAMtools\u003c/em\u003e v 1.3.1 (Li et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). As no reference genome was available for BTF, we used the RefSeq zebra finch (\u003cem\u003eTaeniopygia guttata\u003c/em\u003e) genome assembly \u0026lsquo;bTaeGut1.4.pri\u0026rsquo; (GCF_003957565.2) from NCBI GenBank.\u003c/p\u003e\n\u003ch3\u003eSex inference\u003c/h3\u003e\n\u003cp\u003eThe sex of each bird was inferred using a method adapted from \u003cem\u003eseGMM\u003c/em\u003e (Liu et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and \u003cem\u003eseXY\u003c/em\u003e (Qian et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). First, unfiltered SNPs were mapped against the Z and W chromosomes of the zebra finch (\u003cem\u003eTaeniopygia guttata\u003c/em\u003e, RefSeq assembly GCF_003957565.2) using the function \u0026lsquo;gl.blast\u0026rsquo; in \u003cem\u003edartR\u003c/em\u003e. Mapped sequences were filtered to retain those with percent identity\u0026thinsp;\u0026gt;\u0026thinsp;80%, E-value\u0026thinsp;\u0026lt;\u0026thinsp;1e\u003csup\u003e\u0026minus;\u0026thinsp;10\u003c/sup\u003e, retaining only one SNP per RAD tag, reproducibility\u0026thinsp;\u0026gt;\u0026thinsp;0.98, and removing loci located in pseudoautosomal regions or mapping to both Z and W chromosomes. Three sex-associated metrics were estimated per individual: W missing rate, Z/W ratio and Z heterozygosity. Sex was inferred with the model-based clustering function \u0026lsquo;Mclust\u0026rsquo; from \u003cem\u003emclust\u003c/em\u003e v 6.0.0 (Scrucca et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), with the expectation-maximisation (EM) algorithm to cluster samples into two groups (male or female) (Figure S2).\u003c/p\u003e\n\u003ch3\u003ePopulation genetic analyses\u003c/h3\u003e\n\u003cp\u003eGenetic structure was analysed using four individual-based approaches: 1) a principal coordinates analysis (PCoA) with \u003cem\u003eade4\u003c/em\u003e v 1.7\u0026ndash;22 using Euclidian genotypic distance between samples; 2) a discriminant analysis of principal components (DAPC) with \u003cem\u003eadegenet\u003c/em\u003e v 2.1.10 (Jombart \u0026amp; Ahmed, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2011\u003c/span\u003e); 3) SNMF clustering with \u003cem\u003eLEA\u003c/em\u003e v 3.14.0; and 4) the Bayesian clustering approach implemented in STRUCTURE v 2.3 (Jombart, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) via \u003cem\u003edartR.\u003c/em\u003e For the DAPC, we compared results using \u003cem\u003ea priori\u003c/em\u003e groups and \u003cem\u003ede-novo\u003c/em\u003e clustering, as these approaches can impact inference of genetic structure when population differentiation is low (Miller et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The optimal number of PC\u0026rsquo;s was chosen using the function \u0026lsquo;optim.a.score\u0026rsquo; with 1000 simulations. STRUCTURE settings included no prior site information and used the admixture model with correlated allele frequencies (Falush et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), with a burn-in/MCMC setting 10,000/100,000 and 5 independent runs for each value of \u0026#119870; = 1 to \u0026#119870; = 6. The most likely value of \u0026#119870; was identified using the Evanno method (Evanno et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), and CLUMPP v 1.1.2 to consolidate results from multiple runs (Jakobsson \u0026amp; Rosenberg, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). To assess recent migration and identify individuals with mixed ancestry, STRUCTURE was repeated for the \u0026lsquo;related\u0026rsquo; dataset, using regions (north, central, south) as prior site information, \u0026#119870; = 3, \u0026lsquo;gensback\u0026rsquo; = 3 (i.e., back to great-grandparents) and migration prior\u0026thinsp;=\u0026thinsp;0.05.\u003c/p\u003e \u003cp\u003eSpatial genetic structure was examined using spatial PCA (sPCA) and spatial autocorrelation analysis (SAA). sPCA detects cryptic spatial structure by decomposing genetic variation into variance and spatial autocorrelation (Moran\u0026rsquo;s index, \u003cem\u003eI\u003c/em\u003e) with \u003cem\u003eadegenet\u003c/em\u003e. sPCA was run using an inverse distance matrix for weighting connectivity in the neighbour network (type 7), and significant global (high variance and positive \u003cem\u003eI\u003c/em\u003e) and local structures (high variance and negative \u003cem\u003eI\u003c/em\u003e) were detected with permutation tests (999 replications). SAA was performed with \u003cem\u003edartR\u003c/em\u003e, following Smouse and Peakall (\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e1999\u003c/span\u003e), using pairwise genotypic correlations (\u003cem\u003er\u003c/em\u003e) based on Euclidian genotypic distance, uneven distance classes to optimize sample sizes, 9999 bootstraps to compute 95% C.I\u0026rsquo;s around mean \u003cem\u003er\u003c/em\u003e within distance classes and 9999 permutations to test the null hypothesis of no spatial structure (i.e., random mating).\u003c/p\u003e \u003cp\u003eSex-biased dispersal was assessed using kinship permutation tests and the sex-biased dispersal tests described in Goudet et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Mean kinship was compared between sexes, localities and regions using the KING-derived kinship coefficients with GENESIS, accounting for genetic structure. Two-sided permutation tests owing to the non-independence of pairwise kinship data, by reshuffling group labels among samples without replacement (999 permutations). Goudet\u0026rsquo;s sex-biased dispersal tests were implemented with the function \u0026lsquo;sexbias.test\u0026rsquo; in the R package \u003cem\u003eHierfstat\u003c/em\u003e (Goudet, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), using two-sided permutation tests to assess whether four statistics were significantly different between males and females; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e, mean corrected assignment index (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:mAIc\\)\u003c/span\u003e\u003c/span\u003e) and variance (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:vAIc\\)\u003c/span\u003e\u003c/span\u003e). Sex was reshuffled for each permutation, keeping sex ratios and individual locations (site or region) constant.\u003c/p\u003e \u003cp\u003eGlobal \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e, and pairwise \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e between localities or regions were estimated using \u003cem\u003ediveRsity\u003c/em\u003e v 1.9.90 (Keenan et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and \u003cem\u003edartR\u003c/em\u003e (95% C.I.\u0026rsquo;s computed with 999 bootstraps). Genetic diversity metrics were calculated per region (North, Central, South) with \u003cem\u003ediversity\u003c/em\u003e, including allelic richness (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Ar\\)\u003c/span\u003e\u003c/span\u003e), observed heterozygosity (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{H}_{o}\\)\u003c/span\u003e\u003c/span\u003e), unbiased expected heterozygosity (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{H}_{e}\\)\u003c/span\u003e\u003c/span\u003e), and inbreeding coefficient (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eLandscape resistance modelling\u003c/h3\u003e\n\u003cp\u003eTo identify potential barriers to gene flow between BTFS populations in the Desert Uplands Bioregion, we first mapped estimated effective migration surfaces (EEMS) across the study area using a 15km buffer around all sampling locations. EEMS infers deviation from continuous gene flow across the landscape to identify areas of higher or lower than average historic migration (Petkova et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). In brief, EEMS creates a triangular grid (based on the number of subpopulations, or demes) overlayed on the study area, and estimates the expected genetic dissimilarity between pairs of samples assuming a stepping-stone model of migration between demes. Pairwise resistance distances between samples are calculated using a circuit theory-based approach and the estimated effective migration rates are interpolated across geographic space. As the number of demes chosen affects the grid resolution, we tested a range of deme sizes from n\u0026thinsp;=\u0026thinsp;50 to n\u0026thinsp;=\u0026thinsp;500. EEMS was run with default parameters (MCMC iterations\u0026thinsp;=\u0026thinsp;8,000,000, burn-in =\u0026thinsp;1,000,000, thinning\u0026thinsp;=\u0026thinsp;9999), with three independent Markov Chain Monte Carlo (MCMC) chains to assess convergence and average results across runs. Effective migration rates were mapped across the landscape using the R package \u003cem\u003ereemsplot2\u003c/em\u003e v 0.1.0 (Petkova, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFunctional connectivity across the landscape was further explored using the R package \u003cem\u003eResistanceGA\u003c/em\u003e v 4.2.10 (Peterman, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) to test for isolation by resistance (IBR). \u003cem\u003eResistanceGA\u003c/em\u003e first optimizes individual landscape resistance surfaces by fitting a linear mixed effects model with a maximum likelihood population effects parameterization (MPLE) to the data, with pairwise genetic distance as the response variable and pairwise effective distance as the predictor variable. \u003cem\u003eResistanceGA\u003c/em\u003e then uses a circuit theory-based approach to measure current flow between individuals across each resistance surface, using the package \u003cem\u003eCircuitscape\u003c/em\u003e v 5.13.1 (Hall et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) within \u003cem\u003eJulia\u003c/em\u003e v 1.10.0 (Bezanson et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe assessed the impact of seven landscape variables on inter-individual genetic distance (calculated with \u0026lsquo;gl.dist.ind\u0026rsquo; in \u003cem\u003edartR\u003c/em\u003e with method\u0026thinsp;=\u0026thinsp;Euclidian). Previous studies indicate that vegetation density and structure, as well as distance to permanent water sources, influence BTFS dispersal and flock sizes (Rechetelo, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Rechetelo et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and influence genetic structure in the Townsville population (Tang, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tang et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In this study, we tested three categorical and four continuous landscape variables that captured vegetation structure, habitat suitability and water availability (Tables\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, S1, Figures S3-4). Details on the preparation of these landscape raster layers are provided in the Electronic Supplementary Material. Roads and tracks were excluded from the analysis because the study area is intersected by only one public road which receives a low volume of traffic, and BTFS have been observed crossing this road as well as nesting and foraging along the edge (Brad Dreis \u003cem\u003epers. comm.\u003c/em\u003e).\u003c/p\u003e \u003cp\u003eFirst, each landscape variable was optimized independently using the functions \u0026lsquo;GA.prep\u0026rsquo;, \u0026lsquo;JL.prep\u0026rsquo; and \u0026lsquo;SS.optim\u0026rsquo; in \u003cem\u003eResistanceGA\u003c/em\u003e. For each continuous resistance surface, we tested all eight data transformations available in \u003cem\u003eResistanceGA\u003c/em\u003e to choose the optimal transformation. The function \u0026lsquo;JL.prep\u0026rsquo; was used to generate pairwise resistance matrices via \u003cem\u003eCircuitscape\u003c/em\u003e in Julia. \u0026lsquo;SS.optim\u0026rsquo; carried out resistance surface optimization for each layer, with two independent runs (random seed set for each run), as well as for a \u003cem\u003egeographic distance\u003c/em\u003e only model and a \u003cem\u003enull\u003c/em\u003e model (intercept only). We used the Akaike information criteria corrected for small sample sizes (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e) from the fitted MLPE model to assess the support of each resistance surface model (Beninde et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Shirk et al., \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). \u0026lsquo;Pseudo\u0026rsquo; bootstrapping was carried out for the best run (lowest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e) for each surface using the function \u0026lsquo;Resist.boot\u0026rsquo;. Correlations between landscape layers (both pre- and post-optimization) were calculated using Pearson correlation coefficients and the function \u0026lsquo;layerStats\u0026rsquo; from the R package \u003cem\u003eraster\u003c/em\u003e v 3.6\u0026ndash;26.\u003c/p\u003e \u003cp\u003eTo test whether models based on only one or multiple landscape variables were better supported, and to assess functional connectivity over the entire landscape, we created optimized multi-feature resistance surfaces (OMFRS) using an iterative approach. We only included models that significantly impacted functional connectivity more than \u003cem\u003egeographic distance\u003c/em\u003e alone, i.e., those with a bootstrapped \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e score\u0026thinsp;\u0026lt;\u0026thinsp;2 or more below the \u003cem\u003egeographic distance\u003c/em\u003e only model in both runs. We used the function \u0026lsquo;MS.optim\u0026rsquo; in \u003cem\u003eResistanceGA\u003c/em\u003e, excluding one of each strongly correlated optimized resistance surfaces (r\u0026thinsp;\u0026gt;\u0026thinsp;0.6, retaining the surface with the smallest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e). This function iteratively adds optimized single resistance surfaces to a combined model, starting with the surface with the lowest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e, adding the surface with the next lowest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e, and repeating the optimization and bootstrapping procedure. Each single surface layer added sequentially to the combined model was only retained if the bootstrapped \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e of the new combined model was lower than that of the prior model. Lastly, we mapped the least-cost resistance pathways between samples with \u003cem\u003eCircuitscape\u003c/em\u003e, for each optimized resistance surface (best run based on lowest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eContemporary effective population size\u003c/h2\u003e \u003cp\u003eContemporary \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e was estimated using the bias-corrected linkage disequilibrium (LD) method with \u003cem\u003eNeEstimator\u003c/em\u003e v 2.1 (Waples \u0026amp; Do, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{crit}\\)\u003c/span\u003e\u003c/span\u003e = 0.02 and calculating bias-corrected jack-knife 95% confidence intervals (Waples and Do \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Given the presence of overlapping generations for BTF, this raw estimate was interpreted as the effective number of breeders in the parent generation which produced the cohorts sampled (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Raw{\\widehat{N}}_{b}\\)\u003c/span\u003e\u003c/span\u003e). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Raw{\\widehat{N}}_{b}\\)\u003c/span\u003e\u003c/span\u003e was adjusted to account for this bias and converted to \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e using the equations in Table S2, and two life history traits (age at sexual maturity (α) and adult lifespan (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:AL\\)\u003c/span\u003e\u003c/span\u003e)). To account for the limited ecological knowledge available for BTF, we converted \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e using α\u0026thinsp;=\u0026thinsp;0.5 or 1 year, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:AL\\)\u003c/span\u003e\u003c/span\u003e = 4 to 8 years (Shephard, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e1989\u003c/span\u003e). No adjustments were made for physical linkage as we assumed the filtered SNP data were physically unlinked (retaining one SNP per RAD tag). As LD is built up over generations, we interpreted \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e as the short-term harmonic mean across a few recent generations. We also estimated \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e separately for the North-Central (n = 87) and South (n = 20) groups. To account for differences in sample size, we generated 100 bootstrap replicates for the North-Central group by randomly subsampling 20 individuals each time (with replacement) and recalculating \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDemographic history\u003c/h3\u003e\n\u003cp\u003eDemographic history was inferred using two site frequency spectrum (SFS) based approaches; Stairway Plot v 2.0 (Liu \u0026amp; Fu, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and δaδi v 2.3 (Gutenkunst et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). For both approaches, 1-dimensional SFS\u0026rsquo;s were generated with ANGSD v 0.940, using the genome-mapped BAM files and applying the following filters: minimum base quality score\u0026thinsp;\u0026gt;\u0026thinsp;20, minimum mapping quality\u0026thinsp;\u0026gt;\u0026thinsp;20, individual read depth\u0026thinsp;\u0026gt;\u0026thinsp;2 (in at least 50% of samples) and \u0026lt;\u0026thinsp;1000, doSaf\u0026thinsp;=\u0026thinsp;1 to calculate allele frequency likelihoods based on individual genotype likelihoods (GL) obtained using \u003cem\u003eSAMtools\u003c/em\u003e (GL\u0026thinsp;=\u0026thinsp;1). Folded SFS\u0026rsquo;s were generated as we lacked information on ancestral allelic states.\u003c/p\u003e \u003cp\u003eStairway Plot was run with default settings, with sequence length (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:L\\)\u003c/span\u003e\u003c/span\u003e) calculated as the total number of variant and invariant sites after filtering. Demographic inference with δaδi was performed with \u003cem\u003edadi-cli\u003c/em\u003e (Huang et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Five 1D models were tested, described in Table S3. Parameters for each model were optimized using the function \u0026lsquo;InferDM\u0026rsquo; with five independent runs each with 5000 optimizations and the --global-optimization flag, until convergence was reached (\u0026lsquo;BestFits\u0026rsquo;). As loci were considered \u0026lsquo;linked\u0026rsquo;, models were compared using model scores (best model yields a score of 1) and Akaike weights (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{W}_{AIC}\\)\u003c/span\u003e\u003c/span\u003e) (evaluates the relative probability of each model, as outlined in Rougeux et al. (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) (equations in Table S4). 95% confidence intervals around parameter estimates were obtained with 100 bootstrapped SFS\u0026rsquo;s generated with easySFS (with replacement) and the \u0026lsquo;StatDM\u0026rsquo; function in \u003cem\u003edadi-cli\u003c/em\u003e. Parameters for the best-fit model were converted into measures of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e and time in years, using the equations listed in Table S5. As the germline mutation rate (per site per generation, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mu\\:}_{G}\\)\u003c/span\u003e\u003c/span\u003e) is unknown for BTF, conversions used the estimate for zebra finch, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mu\\:}_{G}\\)\u003c/span\u003e\u003c/span\u003e = 5.85\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\times\\:\\)\u003c/span\u003e\u003c/span\u003e10\u003csup\u003e\u0026minus;9\u003c/sup\u003e (Bergeron et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and a generation length (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:g\\)\u003c/span\u003e\u003c/span\u003e) of 2 years (Bird et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). To account for the limited ecological information available for BTF, we also used a longer generation length of 3.5 years (Garnett et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eAfter filtering, we retained 14,788 neutral SNPs (107 samples) for analyses of neutral genetic structure and diversity, excluding highly related individuals (full-sibs and half-sibs, n\u0026thinsp;=\u0026thinsp;48) (Table S6). Sex inference identified 70 females and 47 males in the \u0026lsquo;unrelated\u0026rsquo; dataset (81 F and 74 M total). For analyses of historical demography, we retained 4,438,791 sites (including 80,382 variant sites) after filtering for Stairway and δaδi analyses (all samples pooled).\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eEvidence of fine-scale genetic structure\u003c/h2\u003e \u003cp\u003eUsing the \u0026lsquo;unrelated\u0026rsquo; dataset, global \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e indicated weak but significant genetic structure (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e = 0.011). SNMF clustering with LEA provided support for a single ancestral population, with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:K\\)\u003c/span\u003e\u003c/span\u003e = 1 yielding the lowest cross-entropy. The DAPC, STRUCTURE and PCoA analyses all suggested \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:K\\)\u003c/span\u003e\u003c/span\u003e \u0026gt; 1, and consistently clustered most South samples separately from Central and North samples (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, S5-6). Pairwise \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e was significantly different from 0 between all regions but was highest between North and South (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e = 0.018, p \u0026lt; 0.001), followed by Central and South (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e = 0.0136, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and North and Central (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e = 0.0057, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). At the site level, pairwise \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e ranged from 0.005 (p \u0026gt; 0.05) to 0.025 p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and was highest between localities in the South (CMB or LFT) and all other localities in the Central and North regions (p \u0026lt; 0.001 for all comparisons) (Figure S7). Measures of genetic diversity were similar between regions, although \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Ar\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e were both significantly higher in the Central sub-population (Table S7). Global \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e was 0.160.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eBased on assignment tests in STRUCTURE with prior population information, most samples from the Central region (97.9%) were correctly assigned to this region with high probability (P\u0026thinsp;\u0026gt;\u0026thinsp;0.9), and none were identified as migrants from another region (Table S8). Most northern samples (81.5%) also assigned to the central region, with only 7.4% (n\u0026thinsp;=\u0026thinsp;2) assigning to the north with no mixed ancestry and 11.1% (n\u0026thinsp;=\u0026thinsp;3) having recent ancestry in the central region (back to grandparents) (P\u0026thinsp;\u0026gt;\u0026thinsp;0.9, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). In the southern region, only 22.6% of samples were correctly assigned to this region with no mixed ancestry, 29.0% were identified as migrants from the central region (n\u0026thinsp;=\u0026thinsp;9) and 16.1% (n\u0026thinsp;=\u0026thinsp;5) had recent ancestry in the central region (grandparents) (P\u0026thinsp;\u0026gt;\u0026thinsp;0.9, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Further, after updating KING kinship coefficients to account for genetic structure, we detected 45 pairs of first-degree relatives (full-sibs or parent-offspring, 0.375\u0026thinsp;\u0026gt;\u0026thinsp;\u003cem\u003ek\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.1875) across 48 samples, and 61 pairs of second-degree relatives (half-sibs or grandparent-offspring, 0.1875\u0026thinsp;\u0026gt;\u0026thinsp;\u003cem\u003ek\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.09375) across 74 samples, following the midpoint cut offs in Iacchei et al. (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). First-degree relatives were detected across years and wet/dry seasons and first- or second-degree relatives were always found in the same region (Table S9).\u003c/p\u003e \u003cp\u003eKinship permutations tests indicated significant fidelity, at both local and regional scales, with mean within-site and within-region kinship being significantly higher than between-site and between-region kinship, respectively (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Figure S8, Table S10). The sPCA indicated significant global structure (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), i.e., high variance and positive spatial autocorrelation (Figure S9). SAA revealed a strong signal of isolation by distance (IBD), with mean genotypic correlation \u0026lsquo;\u003cem\u003er\u003c/em\u003e\u0026rsquo; among samples being significantly different from 0 for all distance class bins (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, S10). Mean \u0026lsquo;\u003cem\u003er\u003c/em\u003e\u0026rsquo; was significantly higher than expected under conditions of random mating for the 2.5 km distance class (i.e., within-site comparisons) and significantly lower than expected for distance classes\u0026thinsp;\u0026ge;\u0026thinsp;30 km (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, S10). This pattern was weaker within the central sub-population where kinship was lower overall (Figure S10, Table S10), and we lacked sufficient sites to carry out SAA for northern or southern samples separately. There was no evidence of sex-biased dispersal, based on kinship permutation tests or SAA (Table S10). For Goudet\u0026rsquo;s sex-biased dispersal tests, only mean variance in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:AIc\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{IS}\\)\u003c/span\u003e\u003c/span\u003e were significantly higher among females (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eHabitat suitability influences gene flow across the landscape\u003c/h2\u003e \u003cp\u003eResults of the EEMS analysis were similar between different deme sizes of n\u0026thinsp;=\u0026thinsp;50 to n\u0026thinsp;=\u0026thinsp;500. Using n\u0026thinsp;=\u0026thinsp;100 demes (approximately 1 deme per 2.25 km\u003csup\u003e2\u003c/sup\u003e), the estimated effective migration rate was significantly higher than the average rate of migration within both the Central and South subpopulations, and the migration rate was significantly lower than expected between these two subpopulations (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, S11).\u003c/p\u003e \u003cp\u003eAll single surface models tested with \u003cem\u003eresistanceGA\u003c/em\u003e outperformed the geographic distance only and null models (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varDelta\\:AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e \u0026gt; 2, Tables\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, S11). The best supported single surface model was \u003cem\u003ehabitat suitability\u003c/em\u003e (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e = -33,801.07), with areas of suitable habitat having the lowest resistance to gene flow, i.e., Eucalypt dominated vegetation on alluvial soils, Eucalypt woodland on paleo-alluvial sediments and Eucalypt/\u003cem\u003eAcacia\u003c/em\u003e vegetation on metamorphosed sediments (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The \u003cem\u003ehabitat suitability\u003c/em\u003e model was the best supported single surface model 99.3% of the time (out of 1,000 bootstrap iterations). The next best single surface models were \u003cem\u003ecanopy cover\u003c/em\u003e (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varDelta\\:AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e = 13.7) and \u003cem\u003ewet season distance to water\u003c/em\u003e (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varDelta\\:AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e = 99.6) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For canopy cover, open woodland (6\u0026ndash;11% plant cover fraction) and woodland (11\u0026ndash;30%) had the lowest resistance to gene flow, while scattered trees (\u0026thinsp;\u0026lt;\u0026thinsp;6%) and forest (\u0026thinsp;\u0026gt;\u0026thinsp;30%) had higher resistance. Wet season distance to water had better support than the dry season model, and a similar pattern was observed for the bare earth models (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), indicating that the availability of water and vegetation during the wet season had a stronger influence on gene flow relative to the dry season.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBest run results of the single resistance surface optimization procedure implemented by \u003cem\u003eResistanceGA\u003c/em\u003e for black-throated finch subpopulations around the Carmichael Coal Mine. Single surface models are listed in order of increasing \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e is the average \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e value obtained for each model in 1,000 bootstrap iterations; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:k\\)\u003c/span\u003e\u003c/span\u003e is the number of parameters; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varDelta\\:AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e is the difference in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e between the top model (lowest \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e) and each subsequent model; and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{mR}^{2}\\)\u003c/span\u003e\u003c/span\u003e is the average marginal \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e of 1,000 bootstrap iterations.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePredictor\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBest Run\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{k}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{A}\\varvec{I}\\varvec{C}}_{\\varvec{c}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{\\varDelta\\:}\\varvec{A}\\varvec{I}\\varvec{C}}_{\\varvec{c}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{m}\\varvec{R}}^{2}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabitat suitability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33801.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.170\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanopy cover\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33787.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e13.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.167\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet season distance to water\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33701.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e99.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.233\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWet season bare earth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33700.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e100.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.242\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry season bare earth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33674.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e126.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.217\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCanopy height\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33648.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e153.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry season distance to water\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33638.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e162.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.159\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGeographic distance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33613.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e187.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNull model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-33035.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e765.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe multi-surface resistance model with the strongest support was a combination of \u003cem\u003ehabitat suitability\u003c/em\u003e and \u003cem\u003ewet season distance to water\u003c/em\u003e (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e = -18,649.5, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{W}_{AIC}\\)\u003c/span\u003e\u003c/span\u003e = 1.00). However, this model was only the top model 28.2% of the time out of 1,000 pseudo-bootstrap iterations (Table S12), with the single surface model \u003cem\u003ehabitat suitability\u003c/em\u003e ranking as the top model 53.1% of the time. In all multi-surface models tested, \u003cem\u003ehabitat suitability\u003c/em\u003e was identified as the most important variable, contributing \u0026gt; 50% to each combined model. The next best multi-surface model was \u003cem\u003ehabitat suitability*wet season distance to water*wet season bare earth\u003c/em\u003e (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varDelta\\:AIC}_{c}\\)\u003c/span\u003e\u003c/span\u003e = 4.18) (Table S12).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eRelatively small contemporary \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{N}}_{\\varvec{e}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/h2\u003e \u003cp\u003eMean estimates of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e across several recent generations ranged from 1,150 to 1,440 based on a \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{crit}\\)\u003c/span\u003e\u003c/span\u003e of 0.02 and depending on the life history traits used for bias corrections (lower 95% C.I. ranged from 940 to 1,180, upper 95% C.I. 1,470 to 1,840) (Table S13). When considering South and North-Central groups as two distinct genetic units, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e for the South group (n\u0026thinsp;=\u0026thinsp;20) ranged from 290 to 360 (lower 95% C.I. from 210 to 260; upper 95% C.I. from 490 to 610) for \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{crit}\\)\u003c/span\u003e\u003c/span\u003e = 0.03, based on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{crit}\\)\u003c/span\u003e\u003c/span\u003e rule of thumb proposed by Waples and Do (\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) for n \u0026le; 25 (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1/2n\u0026lt;{P}_{crit}\u0026lt;1/n\\)\u003c/span\u003e\u003c/span\u003e). For the North-Central group (n = 87), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e ranged from 1,280 to 1,610 (lower 95% C.I. 1,010 to 1,270, upper 95% C.I. 1,750 to 2,200; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{crit}\\)\u003c/span\u003e\u003c/span\u003e = 0.02). Using n = 20, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e for the North-Central group was at least twice as large as the South in 90% of bootstrap replicates. Across all life-history traits tested, the average bootstrapped \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e for the North-Central subpopulation ranged from 970 to 1,220.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eEvidence of two historical genetic bottlenecks\u003c/h2\u003e \u003cp\u003eWe carried out demographic history analysis for all samples pooled together (\u0026lsquo;unrelated\u0026rsquo; dataset, subset of 20 samples per region). Although we found evidence of fine-scale genetic structure, we assumed the population was panmictic overall with respect to allele frequencies, given the relatively low global \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e and pairwise \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{ST}\\)\u003c/span\u003e\u003c/span\u003e between regions (\u0026lt;\u0026thinsp;0.02), suggesting drift connectivity (Lowe \u0026amp; Allendorf, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). The Stairway plot analysis inferred a historical population expansion beginning\u0026thinsp;\u0026asymp;\u0026thinsp;367,000 years ago and lasting\u0026thinsp;\u0026asymp;\u0026thinsp;127,000 years (2\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\times\\:\\)\u003c/span\u003e\u003c/span\u003e increase from ancestral \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e stabilized until a bottleneck beginning\u0026thinsp;\u0026asymp;\u0026thinsp;700 years ago (82% decrease) followed by a more recent bottleneck\u0026thinsp;\u0026asymp;\u0026thinsp;100 years ago (further 63% decrease from the first bottleneck) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). With a longer generation length of 3.5 years, the expansion period detected by Stairway began\u0026thinsp;\u0026asymp;\u0026thinsp;524,000 years ago, the 1st bottleneck\u0026thinsp;\u0026asymp;\u0026thinsp;1,600 years ago and the 2nd bottleneck\u0026thinsp;\u0026asymp;\u0026thinsp;200 years ago.\u003c/p\u003e \u003cp\u003eThe best fit model in δaδi was the 3 epoch model (3EM, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varDelta\\:AIC\\)\u003c/span\u003e\u003c/span\u003e \u0026gt; 10, Table S14) which inferred a historical population expansion beginning \u0026asymp; 321,000 years ago (95% C.I. = 103,000\u0026ndash;539,000) to around 2\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\times\\:\\)\u003c/span\u003e\u003c/span\u003e the ancestral \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e, and a recent bottleneck \u0026asymp; 60 years ago (95% C.I. = 50\u0026ndash;70) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, Table S15). δaδi detected the most recent population decline compared to Stairway, with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e dropping to \u0026asymp; 900 within the last 30 generations, similarly low to the estimate by \u003cem\u003eNeEstimator\u003c/em\u003e (1,190), which reflects the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e of the past few generations. With \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:g\\)\u003c/span\u003e\u003c/span\u003e = 3.5 years, the expansion began \u0026asymp; 562,000 years ago and the bottleneck \u0026asymp; 100 years ago.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe endemic and endangered southern black-throated finch has experienced severe range contractions and habitat fragmentation due to anthropogenic processes such as agriculture and land clearing (Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In this study, we assessed genetic connectivity in one of the last remaining stronghold populations in the Desert Uplands Bioregion. We found genetic structuring over relatively short distances and limited effective dispersal that was predominantly influenced by the availability of suitable habitat. We also detected the genetic signatures of two historical bottlenecks and relatively small contemporary effective population sizes (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e near or below 1000), suggesting BTFS are at risk of losing genetic diversity and adaptive potential in the future.\u003c/p\u003e \u003cp\u003eBTFS exhibited restricted recent gene flow over relatively short distances (16\u0026ndash;45 km) and lower effective migration between the North-Central and South regions. Assignment tests indicated mixed ancestry within the South region, unidirectional gene flow from Central to South, and no recent gene flow between North and South regions. Additionally, we found evidence of IBD and regional philopatry, and no first- or second-degree relatives were recorded between finches in different regions. These findings are consistent with previous studies on black-throated finches, as discussed below. BTF are philopatric, known to return to the same nest over multiple years, and juveniles remain within family groups for several months after fledging which may promote site fidelity through social learning (Higgins, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Isles, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; NRA, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Zann, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e1976\u003c/span\u003e). Radio tracking data suggest BTFS in the Desert Uplands population are generally sedentary over short timescales, dispersing up to 4.5 km and maintaining stationary home ranges over a 1 month period (van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In the Townsville population, restricted gene flow has similarly been observed over relatively short distances (10\u0026ndash;20 km) (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Tang, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tang et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). While data on long-distance dispersal in BTFS is limited, dietary diversification may be favoured over long-distance dispersal in response to changes in resource availability (Rechetelo et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; van Osta et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Limited effective dispersal may have implications for the long-term persistence of the Desert Uplands stronghold population, and for the broader population of BTFS across Queensland.\u003c/p\u003e \u003cp\u003eVegetation communities and water availability likely play a larger role in restricting gene flow than geographic distance. Habitat suitability was the strongest predictor of genetic connectivity across the study area. The habitat separating the North-Central and South regions was characterized by highly fragmented suitable habitat and very low (\u0026lt;\u0026thinsp;6%) or very high canopy cover (\u0026gt;\u0026thinsp;30%). Studies on the Townsville population indicate that the presence of a large waterbody (the Ross River Dam), urbanization, vegetation structure and elevation are key barriers to gene flow among BTFS subpopulations (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Tang, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Compared to the Townsville population, subpopulations of Desert Uplands finches are not separated by large or permanent waterbodies, and thus vegetation structure and habitat suitability may be more influential in restricting gene flow in this population. In other granivorous birds, landscape features such as reduced forest cover, increased agricultural land or water barriers are linked to restricted gene flow, including the golden-cheeked warbler (\u003cem\u003eDendroica chrysoparia\u003c/em\u003e) (Lindsay et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) and song sparrow (\u003cem\u003eMelospiza melodia\u003c/em\u003e) (Wilson et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In the black-capped chickadee (\u003cem\u003ePoecile atricapillus\u003c/em\u003e), a woodland bird species with a high dispersal propensity, spatial patterns of genetic variation were influenced by small breaks in riparian woodlands that function as dispersal corridors across unsuitable grassland habitat (Adams \u0026amp; Burg, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Similarly, for the Desert Uplands BTFS population, the fragmentation of suitable woodland habitat between the North-Central and South regions appears to impede effective dispersal across larger areas of unsuitable grassland habitat. Across the range of the southern black-throated finch, there may be subtle differences between populations in dispersal ecology and landscape barriers that restrict gene flow over relatively short distances.\u003c/p\u003e \u003cp\u003eEstimates of contemporary \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e over a few recent generations were low in the Desert Uplands population of BTFS. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e informs the risk of losing genetic diversity due to drift over generations (Luikart et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Tallmon et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). The \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e estimates for BTFS are near or below the recommended minimum level of 1,000 needed to maintain evolutionary potential in the long term (Frankham et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), particularly in the southern region (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e ~ 300). Adult census sizes (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{c}\\)\u003c/span\u003e\u003c/span\u003e, total number of potential breeders) are poorly known for BTFS. The most recent estimate of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{c}\\)\u003c/span\u003e\u003c/span\u003e is around 450\u0026ndash;1,000 in the largest population (Desert Uplands Bioregion), and 633\u0026ndash;1,400 across their range, but these estimates are classified as having \u0026lsquo;low\u0026rsquo; reliability (Buosi et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). While there isn\u0026rsquo;t a simple linear relationship between \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{c}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e is often smaller than \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{c}\\)\u003c/span\u003e\u003c/span\u003e, with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}/{N}_{c}\\)\u003c/span\u003e\u003c/span\u003e ranging from 0.28 to 0.40 in other finch species (Frankham, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Overall, this study provides a baseline for continued monitoring of the genetic health of the Desert Uplands population.\u003c/p\u003e \u003cp\u003eThe recent demographic history reconstructed by Stairway and δaδi is consistent with known range contractions for BTFS. The extent of occurrence of BTFS has contracted by 50\u0026ndash;80% since the late 1900\u0026rsquo;s, and likely by over 90% since pre-European times (BTFRT, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2007\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Franklin, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Reside et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This coincides with the recent genetic bottleneck detected by Stairway and δaδi (60\u0026ndash;100 years ago), and the more distant bottleneck detected by Stairway around 700 years ago. Severe declines in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{N}_{e}\\)\u003c/span\u003e\u003c/span\u003e are common in many species around the start of the last glacial period (LGP, 110,000 ya), including in the medium ground finch (\u003cem\u003eGeospiza fortis\u003c/em\u003e) (Nadachowska-Brzyska et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Using WGS data, Hooper et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) detected a long-term decline in BTFS beginning around 11,300 years ago, predating European arrival. The Townsville and Desert Uplands (Galilee Basin) populations experienced asymmetric migration until around 2,000 years ago, when migration is thought to have dropped to negligible levels (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In this study, we inferred a population expansion occurring\u0026thinsp;~\u0026thinsp;300,000 to 500,000 years ago, consistent with expansions occurring in other bird species during the early to mid-Pleistocene period (Nadachowska-Brzyska et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The demographic history of the northern subspecies (BTFN) is characterized by population expansions; and estimates of divergence times between BTF and long-tailed finches (LTF, \u003cem\u003eP. acuticauda\u003c/em\u003e and \u003cem\u003ehecki\u003c/em\u003e) range from 600,000\u0026ndash;1,000,000 years ago (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Jennings \u0026amp; Edwards, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Thus, the expansion event we detected in the Desert Uplands population of BTFS may reflect an earlier signature of population expansion, prior to the genetic isolation of BTFS and BTFN. Collectively, evidence of recent genetic bottlenecks within the Desert Uplands Bioregion and negligible gene flow between the two stronghold populations, suggests the long-term viability of BTFS may be compromised.\u003c/p\u003e \u003cp\u003eThe black-throated finch (northern and southern subspecies) was downlisted to \u0026lsquo;Least Concern\u0026rsquo; on the IUCN Red List in 2016, as the rate of decline did not adhere to Red List criteria (BirdLife International, 2024). Buosi et al. (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) proposed that BTFS may be eligible for a higher classification based on category A of the Red List criteria (population decline projected to be \u0026gt;\u0026thinsp;50\u0026ndash;80% in three generations based on declines in habitat quality) and category D (250\u0026ndash;1,000 mature individuals and a small, restricted distribution). In our genetic study on BTFS in the largest remaining stronghold population, we found evidence of multiple historical genetic bottlenecks, restricted gene flow over relatively short distances likely driven by the fragmentation of suitable habitat, and small contemporary effective population sizes (290\u0026ndash;360 in the South and 1,280 to 1,610 in the North-Central region). Recent evidence obtained with WGS data also suggests the Townsville and Desert Uplands stronghold populations have been genetically isolated for around 2,000 years (Hooper et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). With further rapid declines predicted in the near future, BTFS may warrant a higher and separate sub-species listing on the IUCN Red List (Buosi et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Currently, RL categories and criteria (version 3.1) do not integrate genetic data (IUCN, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), despite the improvements that could be made in assessing extinction risks (Garner et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Incorporating both genetic and ecological data into management plans would improve conservation planning and help prevent continued declines in the last two remaining stronghold populations of the southern black-throated finch.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank Bravus Mining and Resources for funding this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This study was funded by Bravus Mining and Resources.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e The authors declare no conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval:\u0026nbsp;\u003c/strong\u003eAll methods were carried out under AEC ethics approval CA 2020/07/1392, issued by the Queensland Department of Agriculture and Fisheries; research permit WA0025814, issued by the Queensland Department of Environment, Science and Innovation; and Australian Bird and Bat Banding Scheme authority 2832\u0026ndash;01, issued by the Australian Government Department of Climate Change, Energy, the Environment and Water.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate:\u0026nbsp;\u003c/strong\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u0026nbsp;\u003c/strong\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials:\u0026nbsp;\u003c/strong\u003eSupporting methods and results are available in the Electronic Supplementary Material.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u0026nbsp;\u003c/strong\u003eSG, SD, JVO, JM and BD conceived the project. SG and AS supervised the project. BD, JVO and JM secured funding and managed data collection. SD conducted all statistical analyses and wrote the manuscript, with minor edits from SG, AS and JVO. WA assisted with landscape resistance analyses.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAdams RV, Burg TM (2015) Gene flow of a forest-dependent bird across a fragmented landscape. PLoS ONE 10(11). https://doi.org/10.1371/journal.pone.0140938.\u003c/li\u003e\n \u003cli\u003eAllendorf FW, Luikart G (2009) Conservation and the genetics of populations. John Wiley \u0026amp; Sons.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBaldwin M. (1976). Distribution of the black-throated finch. Australian Birds 11:13-14.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBeaumont MA, Balding DJ (2004). Identifying adaptive genetic divergence among populations from genome scans. Mol Ecol 13(4):969-980.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBeninde J, Wittische J, Frantz AC (2024) Quantifying uncertainty in inferences of landscape genetic resistance due to choice of individual‐based genetic distance metric. Mol Ecol Resour 24(1). https://doi.org/10.1111/1755-0998.13831.\u003c/li\u003e\n \u003cli\u003eBergeron LA, Besenbacher S, Zheng J, Li P, Bertelsen MF, Quintard B, Hoffman JI, Li Z, St. Leger J, Shao C (2023) Evolution of the germline mutation rate across vertebrates. Nature 615:285-291.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBezanson J, Edelman A, Karpinski S, Shah VB (2017) Julia: A fresh approach to numerical computing. SIAM Review 59(1):65-98.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBird JP, Martin R, Ak\u0026ccedil;akaya HR, Gilroy J, Burfield IJ, Garnett ST, Symes A, Taylor J, Şekercioğlu \u0026Ccedil;H, \u0026nbsp;Butchart SH (2020) Generation lengths of the world\u0026apos;s birds and their implications for extinction risk. Conserv Biol 34(5):1252-1261.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eBirdLife International (2024) \u003cem\u003ePoephila cincta\u003c/em\u003e. The IUCN Red List of Threatened Species 2024: e.T22719692A253990937. https://dx.doi.org/10.2305/IUCN.UK.2024-2.RLTS.T22719692A253990937.en.\u0026nbsp;Accessed on 30 May 2025.\u003c/li\u003e\n \u003cli\u003eBrond\u0026iacute;zio ES, Settele J, Diaz S, Ngo HT (2019) Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES secretariat p.1148.\u003c/li\u003e\n \u003cli\u003eBTFRT (2007) National recovery plan for the black-throated finch southern subpecies (Poephila cincta cincta). Black Throated Finch Recovery Team, Department of Environment and Climate Change (NSW) and Queensland Parks and Wildlife Service. Report prepared for the Department of the Environment and Water Resources, Canberra. Department of Environment and Climate Change (NSW), Hurstville and Queensland Parks and Wildlife Service, Brisbane. https://www.dcceew.gov.au/sites/default/files/documents/p-cincta.pdf. Accessed 30 May 2025.\u003c/li\u003e\n \u003cli\u003eBTFRT (2020) Conversation action plan summary for the black-throated finch southern subspecies. Black Throated Finch Recovery Team, Department of Environment and Climate Change (NSW) and Queensland Parks and Wildlife Service. https://blackthroatedfinch.org/wp-content/uploads/2024/01/BTF-Conservation-Action-Plan-Summary.pdf. Accessed 30 May 2025.\u003c/li\u003e\n \u003cli\u003eBuosi P, Vanderduys EP, Reside AE (2021) Southern black-throated finch Poephila cincta cincta. In Garnett ST and Baker GB (eds), The Action Plan for Australian Birds 2020 (pp. 786-789). CSIRO Publishing.\u003c/li\u003e\n \u003cli\u003eConomos MP, Thornton T, Gogarten S (2016) GENetic EStimation and inference in structured samples (GENESIS): statistical methods for analyzing genetic data from samples with population structure and/or relatedness. R package version, 2(0.1).\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eEvanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol Ecol 14(8):2611-2620.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eExcoffier L, Lischer HE (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour 10(3):564-567.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eExposito-Alonso M, Booker TR, Czech L, Gillespie L, Hateley S, Kyriazis CC, Lang PL, Leventhal L, Nogues-Bravo D, Pagowski V (2022) Genetic diversity loss in the Anthropocene. Science 377(6613):1431-1435.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFalush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164(4):1567-1587.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFoll M, Gaggiotti OE (2008) A genome scan method to identify selected loci appropriate for both dominant and codominant markers: a Bayesian perspective. Genetics 180(2):977\u0026ndash;993.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eForshaw JM, Shephard M (2012) Grassfinches in Australia. CSIRO Publishing.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFrankham R (1995) Effective population size/adult population size ratios in wildlife: a review. Genet Res 66(2):95-107.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFrankham R (2005) Genetics and extinction. Biological Conservation, 126(2), 131-140.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFrankham R, Ballou J, Ralls K, Eldridge M, Dudash M, Fenster C, Lacy R, Sunnucks P (2017) Population fragmentation causes inadequate gene flow and increases extinction risk. Genetic management of fragmented animal and plant populations, 87-112.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFrankham R, Bradshaw CJ, Brook BW (2014) Genetics in conservation management: revised recommendations for the 50/500 rules, Red List criteria and population viability analyses. Biol Conserv 170:56-63.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eFranklin DC (1999) Evidence of disarray amongst granivorous bird assemblages in the savannas of northern Australia, a region of sparse human settlement. Biol Conserv 90(1):53-68.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGarner BA, Hoban S, Luikart G (2020) IUCN Red List and the value of integrating genetics. Conserv Gen 21(5):795-801. https://doi.org/10.1007/s10592-020-01301-6.\u003c/li\u003e\n \u003cli\u003eGarnett ST, Szabo JK, Duston G (2011) The Action Plan for Australian Birds 2010. CSIRO.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGoudet J (2005) Hierfstat, a package for R to compute and test hierarchical F‐statistics. Mol Ecol Notes 5(1):184-186.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGoudet J, Perrin N, Waser P (2002) Tests for sex‐biased dispersal using bi‐parentally inherited genetic markers. Mol Ecol 11(6):1103-1114.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGrice AC, Large T, Kahler CP, Heading R (2023) Surveys of a small population of the Endangered Black-throated Finch \u0026apos;Poephila cincta cincta\u0026apos; in North Queensland. Aust Field Ornith 40:152-164.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGruber B, Unmack PJ, Berry OF, Georges A (2018) dartr: An r package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Mol Ecol Resour 18(3):691-699.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eGutenkunst RN, Hernandez RD, Williamson SH, Bustamante CD (2009) Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data. PLoS genetics 5(10). https://doi.org/10.1371/journal.pgen.1000695.\u003c/li\u003e\n \u003cli\u003eHall KR, Anantharaman R, Landau VA, Clark M, Dickson BG, Jones A, Platt J, Edelman A, Shah VB (2021) Circuitscape in Julia: empowering dynamic approaches to connectivity assessment. Land 10(3).\u0026nbsp;https://doi.org/10.3390/land10030301.\u003c/li\u003e\n \u003cli\u003eHiggins P (2006) Handbook of Australian, New Zealand and Antarctic Birds: Volume 7 Boatbill to Starlings: Part B Dunnock to Starlings. Oxford University Press.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eHoban S, Archer FI, Bertola LD, Bragg JG, Breed MF, Bruford MW, Coleman MA, Ekblom R, Funk WC, Grueber CE (2022) Global genetic diversity status and trends: towards a suite of Essential Biodiversity Variables (EBVs) for genetic composition. Biol Rev 97(4):1511-1538.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eHoban S, Bruford MW, Funk WC, Galbusera P, Griffith MP, Grueber CE, Heuertz M, Hunter ME, Hvilsom C, Stroil BK (2021) Global commitments to conserving and monitoring genetic diversity are now necessary and feasible. BioScience 71(9):964-976.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eHoban S, Campbell CD, da Silva JM, Ekblom R, Funk WC, Garner BA, Godoy JA, Kershaw F, MacDonald AJ, Mergeay J (2021) Genetic diversity is considered important but interpreted narrowly in country reports to the Convention on Biological Diversity: Current actions and indicators are insufficient. Biol Conserv 261. https://doi.org/10.1016/j.biocon.2021.109233.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eHolderegger R, Balkenhol N, Bolliger J, Engler JO, Gugerli F, Hochkirch A, Nowak C, Segelbacher G, Widmer A, Zachos FE (2019) Conservation genetics: linking science with practice. Mol Ecol 28(17):3848-3856.\u003c/li\u003e\n \u003cli\u003eHooper DM, Lopez KA, Butcher BG, Lovette IJ, Griffith SC (2025) Recent divergence and microgeographic genetic structure in an endangered Australian songbird: the southern black-throated finch. bioRxiv. https://doi.org/10.1101/2025.04.17.649380.\u003c/li\u003e\n \u003cli\u003eHuang X, Struck TJ, Davey SW, Gutenkunst RN (2023) dadi-cli: Automated and distributed population genetic model inference from allele frequency spectra. bioRxiv. https://doi.org/10.1101/2023.06.15.545182.\u003c/li\u003e\n \u003cli\u003eIacchei M, Ben‐Horin T, Selkoe KA, Bird CE, Garc\u0026iacute;a‐Rodr\u0026iacute;guez FJ, Toonen RJ (2013) Combined analyses of kinship and FST suggest potential drivers of chaotic genetic patchiness in high gene‐flow populations. Mol Ecol, 22(13):3476-3494.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eIsles NL (2007) Patterns of Foraging Behaviour of the Blackthroated Finch during Breeding Season. Dissertation, James Cook University, Townsville, Queensland.\u003c/li\u003e\n \u003cli\u003eIUCN (2012) IUCN Red List Categories and Criteria: Version 3.1. Second edition. International Union for Conservation of Nature and Natural Resources.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eJakobsson M, Rosenberg NA (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinf 23(14):1801-1806.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eJennings WB, Edwards SV (2005) Speciational history of Australian grass finches (Poephila) inferred from thirty gene trees. Evolution 59(9):2033-2047.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eJombart T (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinf 24(11):1403-1405.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eJombart T, Ahmed I (2011) adegenet 1.3-1: new tools for the analysis of genome-wide SNP data. Bioinf 27(21):3070-3071.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eKeenan K, McGinnity P, Cross TF, Crozier WW, Prod\u0026ouml;hl PA (2013) diveRsity: an R package for the estimation and exploration of population genetics parameters and their associated errors. Methods Ecol Evol 4(8):782-788.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eKilian A, Wenzl P, Huttner E, Carling J, Xia L, Blois H, Caig V, Heller-Uszynska K, Jaccoud D, Hopper C (2012) Diversity arrays technology: a generic genome profiling technology on open platforms. In Pompanon F \u0026amp; Bonin A (eds) Data Production and Analysis in Population Genomics. Humana Press, pp 67-89.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLi H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, Subgroup GPDP (2009) The sequence alignment/map format and SAMtools. Bioinf 25(16):2078-2079.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLindsay DL, Barr KR, Lance RF, Tweddale SA, Hayden TJ, Leberg PL (2008) Habitat fragmentation and genetic diversity of an endangered, migratory songbird, the golden-cheeked warbler (Dendroica chrysoparia). Mol Ecol 17(9):2122-2133. https://doi.org/10.1111/j.1365-294X.2008.03673.x\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLiu S, Zeng Y, Wang C, Zhang Q, Chen M, Wang X, Wang L, Lu Y, Guo H, Bu F (2022) seGMM: A New Tool for Gender Determination From Massively Parallel Sequencing Data. Front Genet 13. https://doi.org/10.3389/fgene.2022.850804.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLiu X, Fu YX (2015) Exploring population size changes using SNP frequency spectra. Nat Genet 47(5):555-559.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLowe WH, Allendorf FW (2010) What can genetics tell us about population connectivity? Mol Ecol 19(15):3038-3051.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eLuikart G, Ryman N, Tallmon DA, Schwartz MK, Allendorf FW (2010) Estimation of census and effective population sizes: the increasing usefulness of DNA-based approaches. Conserv Genet 11:355-373.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eMiller JM, Cullingham CI, Peery RM (2020) The influence of a priori grouping on inference of genetic clusters: simulation study and literature review of the DAPC method. Heredity 125(5):269-280.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eMula Laguna J, Reside AE, Kutt A, Grice AC, Buosi P, Vanderduys EP, Taylor M, Schwarzkopf L (2019) Conserving the endangered Black-throated Finch southern subspecies: what do we need to know? Emu - Austral Ornithol 119(4):331-345.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eNadachowska-Brzyska K, Li C, Smeds L, Zhang G, Ellegren H (2015) Temporal dynamics of avian populations during Pleistocene revealed by whole-genome sequences. Curr Biol 25(10):1375-1380.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eNRA (2005) \u0026apos;Enertrade North Queensland Gas Pipeline - Black-throated finch studies (post-construction)\u0026apos; Report prepared by NRA Environmental Consultants to Enertrade.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003ePeterman WE (2018) ResistanceGA: An R package for the optimization of resistance surfaces using genetic algorithms. Methods Ecol Evol 9(6):1638-1647.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003ePetkova D (2024) reemsplots2: Generate plots to inspect and visualize the results of EEMS. R package version 0.1.0, commit fbc2de644d9f591c91b795489b5026f9851f7fd8, \u0026lt;https://github.com/dipetkov/reemsplots2\u0026gt;.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003ePetkova D, Novembre J, Stephens M (2016) Visualizing spatial population structure with estimated effective migration surfaces. Nat Genet 48(1):94-100.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003ePierson JC, Coates DJ, Oostermeijer JGB, Beissinger SR, Bragg JG, Sunnucks P, Schumaker NH, Young AG (2016) Genetic factors in threatened species recovery plans on three continents. Front Ecol Environ 14(8):433-440.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eQian DC, Busam JA, Xiao X, O\u0026rsquo;Mara TA, Eeles RA, Schumacher FR, Phelan CM, Amos CI (2017) seXY: a tool for sex inference from genotype arrays. Bioinf 33(4):561-563.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eRaffard A, Santoul F, Cucherousset J, Blanchet S (2019) The community and ecosystem consequences of intraspecific diversity: A meta‐analysis. Biol Rev 94(2):648-661.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eRechetelo J (2015) Movement, habitat requirements, nesting and foraging site selection: a case study of an endangered granivorous bird, the Black-throated finch Poephila cincta cincta in north-eastern Australia. Dissertation, James Cook University.\u003c/li\u003e\n \u003cli\u003eRechetelo J, Grice A, Reside AE, Hardesty BD, Moloney JM (2016) Movement patterns, home range size and habitat selection of an endangered resource tracking species, the black-throated finch (Poephila cincta cincta). PLoS ONE 11(11). doi:0167210.0161371/journal.pone.0167254.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eReside AE, Beher J, Cosgrove AJ, Evans MC, Seabrook L, Silcock JL, Wenger AS, Maron M (2017) Ecological consequences of land clearing and policy reform in Queensland. Pacific Conserv Biol 23(3):219-230.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eReside AE, Cosgrove AJ, Pointon R, Trezise J, Watson JE, Maron M (2019) How to send a finch extinct. Environmental Science \u0026amp; Policy 94:163-173.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eRougeux C, Bernatchez L, Gagnaire PA (2017) Modeling the multiple facets of speciation-with-gene-flow toward inferring the divergence history of lake whitefish species pairs (Coregonus clupeaformis). Genome Biol Evol 9(8):2057-2074.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eSansaloni C, Petroli C, Jaccoud D, Carling J, Detering F, Grattapaglia D, Kilian A (2011) Diversity Arrays Technology (DArT) and next-generation sequencing combined: genome-wide, high throughput, highly informative genotyping for molecular breeding of Eucalyptus. In BMC proceedings (Vol. 5, No. Suppl 7, p. P54). London: BioMed Central.\u003c/li\u003e\n \u003cli\u003eSchodde R, Mason IJ (1999) Directory of Australian birds: passerines: Passerines. CSIRO publishing.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eSchubert M, Lindgreen S, Orlando L (2016) AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res Notes 9:1-7.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eScrucca L, Fop M, Murphy TB, Raftery AE (2016) mclust 5: clustering, classification and density estimation using Gaussian finite mixture models. The R Journal 8(1): p 289.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eShephard M (1989) Aviculture in Australia: Keeping and Breeding Aviary Birds. Black Cockatoo Press.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eShirk A, Landguth E, Cushman S (2017) A comparison of individual‐based genetic distance metrics for landscape genetics. Mol Ecol Resour 17(6):1308-1317.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eSmouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82(5):561-573.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eTallmon DA, Gregovich D, Waples RS, Baker CS, Jackson J, Taylor BL, Archer E, Martien KK, Allendorf FW, Schwartz MK (2010) When are genetic methods useful for estimating contemporary abundance and detecting population trends? Mol Ecol Resour 10(4):684-692.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eTang LS (2016) Conservation genetics of granivorous birds in a heterogeneous landscape: the case of the black-throated finch (Poephila cincta). Dissertation, James Cook University.\u003c/li\u003e\n \u003cli\u003eTang LS, Smith-Keune C, Grice AC, Moloney JM, Hardesty BD (2017) Genetic structure and diversity of the black-throated finch (Poephila cincta) across its current range. Aust J Zool 64(6):375-384.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003evan Osta JM, Dreis B, Grogan LF, Castley JG (2024) Local resource availability drives habitat use by a threatened avian granivore in savanna woodlands. PLoS ONE 19(8). https://doi.org/10.1371/journal.pone.0306842.\u003c/li\u003e\n \u003cli\u003eVanderduys EP, Reside AE, Grice A, Rechetelo J (2016) Addressing potential cumulative impacts of development on threatened species: the case of the endangered black-throated finch. PLoS ONE 11(3). https://doi.org/10.1371/journal.pone.0148485.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWalters AD, Schwartz MK (2020) Population genomics for the management of wild vertebrate populations. In: Population Genomics: Wildlife. Cham: Springer International Publishing, pp 419-436.\u003c/li\u003e\n \u003cli\u003eWaples RS, Do C (2008) LDNE: a program for estimating effective population size from data on linkage disequilibrium. Mol Ecol Resour 8(4):753-756.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWaples RS, Do C (2010) Linkage disequilibrium estimates of contemporary Ne using highly variable genetic markers: a largely untapped resource for applied conservation and evolution. Evol Appl 3(3):244-262.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWhitlock MC, Lotterhos KE (2015) Reliable detection of loci responsible for local adaptation: inference of a null model through trimming the distribution of F ST. Am Nat 186(1):24-36.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWigginton JE, Cutler DJ, Abecasis GR (2005) A note on exact tests of Hardy-Weinberg equilibrium. Am J Hum Genet 76(5):887-893.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWilson AG, Arcese P, Chan YL, Patten MA (2011) Micro-spatial genetic structure in song sparrows (Melospiza melodia). Conserv Genet 12:213-222.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWoinarski J, Catterall C (2004) Historical changes in the bird fauna at Coomooboolaroo, northeastern Australia, from the early years of pastoral settlement (1873) to 1999. Biol Conserv 116(3):379-401.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eZann R (1976) Distribution, status and breeding of Black-throated Finches Poephila cincta in northern Queensland. Emu \u0026ndash; Austral Ornithol 76(4):201-206.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eZheng X, Zheng MX (2013) Package \u0026lsquo;SNPRelate\u0026rsquo;. A package for parallel computing toolset for relatedness and principal component analysis of SNP data.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"conservation-genetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"coge","sideBox":"Learn more about [Conservation Genetics](https://www.springer.com/journal/10592)","snPcode":"10592","submissionUrl":"https://submission.nature.com/new-submission/10592/3","title":"Conservation Genetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Effective population size, kinship, landscape resistance modelling, functional connectivity, historical demography, conservation genetics","lastPublishedDoi":"10.21203/rs.3.rs-6780931/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6780931/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHabitat fragmentation is a key driver of reduced genetic connectivity and loss of genetic variation among populations, elevating the risk of inbreeding depression and reduced adaptive potential. The endemic, nationally endangered southern black-throated finch (\u003cem\u003ePoephila cincta cincta\u003c/em\u003e) has experienced severe range contractions since the rise of pastoralism. Using a panel of genome-wide single nucleotide polymorphisms, we characterised spatial genetic structure for a remaining stronghold population in the Desert Uplands Bioregion of Queensland, Australia. We mapped effective migration surfaces and tested for isolation by resistance to identify potential barriers to gene flow, estimated contemporary effective population sizes and reconstructed the demographic history of this population. We found evidence of restricted gene flow between localities only 16 km apart and strong isolation by geographic distance. Landscape resistance modelling identified areas of suitable woodland habitat that facilitated effective dispersal. More restricted gene flow in the southern range of this population is likely influenced by the fragmentation of suitable vegetation communities. Contemporary effective population sizes were near or below 1000, and we detected two historical population bottlenecks (\u0026gt;\u0026thinsp;50% decline) occurring around 60\u0026ndash;100 and 700 years ago. Given recent evidence that the Desert Uplands population is genetically isolated from the only other stronghold population in Townsville, the results of this study suggest future losses of genetic diversity and adaptive potential may continue without effective management. To improve the long-term persistence of southern-black throated finch across their range, prioritising the conservation and restoration of habitat that promotes genetic connectivity is essential.\u003c/p\u003e","manuscriptTitle":"Recent historical bottlenecks and restricted gene flow in one of the last remaining stronghold populations of the southern black-throated finch (Poephila cincta cincta)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-10 08:16:40","doi":"10.21203/rs.3.rs-6780931/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-24T02:46:08+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"260613953860819885162619717519796814564","date":"2025-06-11T03:59:58+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-09T02:02:49+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-03T02:26:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-03T02:25:00+00:00","index":"","fulltext":""},{"type":"submitted","content":"Conservation Genetics","date":"2025-05-30T04:33:34+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"conservation-genetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"coge","sideBox":"Learn more about [Conservation Genetics](https://www.springer.com/journal/10592)","snPcode":"10592","submissionUrl":"https://submission.nature.com/new-submission/10592/3","title":"Conservation Genetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"57030dbb-73ee-4366-ab1c-2626d0f99abc","owner":[],"postedDate":"June 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-02T10:23:27+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-10 08:16:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6780931","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6780931","identity":"rs-6780931","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}