Establishment of a non-invasive age estimation method based on fecal DNA methylation levels in brown bears Running title: Fecal DNA-based Epigenetic Clock in Brown Bears

Establishment of a non-invasive age estimation method based on fecal DNA methylation levels in brown bears Running title: Fecal DNA-based Epigenetic Clock in Brown Bears | 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 Article Establishment of a non-invasive age estimation method based on fecal DNA methylation levels in brown bears Running title: Fecal DNA-based Epigenetic Clock in Brown Bears Satoshi Ohara, Shiori Nakamura, Kyogo Hagino, Yuu Yoshimi, Naoya Matsumoto, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9157516/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Age is important for understanding the life history of wildlife. We previously established an age estimation method for brown bears based on blood DNA methylation levels. However, collecting blood samples requires invasive procedures. In this study, we established a non-invasive age estimation method based on DNA methylation levels using fecal DNA collected from 43 brown bears of known age living in both captivity and the wild. Bisulfite pyrosequencing was performed to determine the methylation levels of fecal DNA, and the best model was constructed based on four cytosine-phosphate-guanine (CpG) sites: one adjacent to DLX5 and three adjacent to SLC12A5 . The mean absolute error after leave-one-out cross-validation was 2.08 years, and the median absolute error was 0.99 years; these results demonstrate high accuracy. Furthermore, a method was implemented for quantifying host DNA copy numbers, demonstrating that analyses tend to fail or exhibit large estimation errors when the amount of host DNA is insufficient. This is a significant advancement compared to previous techniques that relied on fecal DNA methylation levels. Applying this method to field surveys will greatly contribute to ecological research and the development of appropriate conservation and management strategies for bears, while facilitating future epigenetic clock studies on other animals. Biological sciences/Biological techniques Biological sciences/Ecology Earth and environmental sciences/Ecology Biological sciences/Genetics Biological sciences/Molecular biology Biological sciences/Zoology epigenetic clock age estimation brown bear DNA methylation feces wildlife management Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Information regarding the age of wild animals is crucial for conducting ecological research and developing appropriate conservation and management strategies 1 – 3 . Traditionally, the age of mammals has been estimated by counting the annual growth layers in the cementum of the teeth 4 . These layers accumulate with age, and the number of layers is used to estimate chronological age. However, this method has several limitations. First, tooth extraction is required, which is highly invasive in living animals. Furthermore, collecting samples from dead individuals requires specialized skills and considerable effort. Second, the accuracy decreases in older individuals because the cementum layers become increasingly indistinct and difficult to count 5 . Third, this method requires advanced technical skills and the results may vary depending on the observer’s subjective judgment 5 . Fourth, the formation rate of cementum layers can vary depending on climatic and nutritional conditions, necessitating species- or region-specific calibration 6 . Over the last two decades, DNA methylation levels have been used as indicators of age 7 . DNA methylation is an epigenetic mechanism that transfers a methyl group to the C5 position of cytosine to form 5-methylcytosine, which primarily occurs on cytosines located within cytosine–guanine dinucleotide (CpG) sites 8 . This is a major epigenetic mechanism involved in gene regulation and cellular differentiation. DNA methylation regulates gene expression by preventing transcription factors from binding to DNA or by altering chromatin states to inhibit transcription factor binding 9 . Previous studies have revealed that the methylation levels at specific CpG sites increase or decrease directionally with age 10 , and these age-associated changes have been consistently observed in different tissues and mammalian species 11 . Statistical models for age prediction (i.e., epigenetic clocks) have been developed based on age-dependent changes in DNA methylation levels 12 . Epigenetic clock research was initially developed in forensic studies on humans using various biological samples such as blood 13 , muscle 14 , and saliva 15 . In recent years, such studies have been extended to non-human animals. Age estimation models have been established using laboratory animals such as mice ( Mus musculus ) 16 and naked mole-rats ( Heterocephalus glaber ) 17 . Similar research has also been conducted on companion and domestic animals, such as dogs ( Canis familiaris ) 18 , cats ( Felis catus ) 19 , 20 , and cattle ( Bos taurus ) 21 . Moreover, epigenetic clock studies have been performed on various wild species, including humpback whales ( Megaptera novaeangliae ) 22 , bottlenose dolphins ( Tursiops truncatus ) 23 , chimpanzees ( Pan troglodytes ) 24 , wolves ( Canis lupus ) 18 , and roe deer ( Capreolus capreolus ) 25 . Because these methods use biological samples such as blood 18 , 26 , muscle 14 , and skin 22 , invasive procedures such as capture, anesthesia, and dart biopsy are required. However, in the field, such procedures impose a physical burden on animals and pose potential risks to researchers. Therefore, there is a growing need to develop less invasive or non-invasive sampling methods that can be performed safely with minimal impact on animals. Representative less-invasive or non-invasive samples include hair, feces, and urine, among which feces are particularly easy to collect from many animal species. Feces contain DNA derived from the intestinal epithelial cells of the host 27 , and recent studies have demonstrated that methylation information can also be obtained from fecal DNA. Age estimation based on fecal DNA has been reported for several species, including Indo-Pacific bottlenose dolphins ( Tursiops aduncus ) 28 , wild mice 29 , Tsushima leopard cats ( Prionailurus bengalensis euptilurus ) 30 . However, studies using fecal DNA are limited, and the target species remain few. The scarcity of methylation studies using fecal DNA may be attributed to several challenges unique to fecal samples. First, collecting feces from free-ranging wild animals of known age across a wide age range, which is essential for constructing an age estimation model, is extremely difficult. Second, fecal DNA is inferior in both quality and quantity compared to DNA obtained from blood or other tissues, and the proportion of host DNA is even lower because it contains DNA derived from intestinal bacteria and food. Third, fecal DNA is highly affected by freshness and environmental exposure, making it difficult to obtain stable and high-quality DNA 31 , 32 . The purpose of this study was to establish an epigenetic clock in feces collected from brown bears ( Ursus arctos ). Brown bears have a lifespan of approximately 20–30 years 33 . Offspring become independent of their mothers at 1.5–2.5 years of age 34 , and physical growth is completed at approximately 5 years in females and 8 years in males 35 . The minimum age at first reproduction is four years 36 . Hibernation lasts for three to seven months 37 , and pregnant females give birth during hibernation between late January and early February 38 . Individual and seasonal variations in body size make it almost impossible to identify the age of bears by appearance, except for cubs-of-the-year 35 , 39 . Therefore, tooth-based methods have traditionally been used to estimate the age of brown bears 36 . Brown bears have low reproductive rates and are vulnerable to overharvesting; thus, population recovery after a decline requires many years 40 . Human–bear conflicts, such as crop depredation, intrusion into residential areas, and attacks on livestock and humans, have become serious global issues 41 , 42 . Therefore, understanding the age structure of brown bear populations is essential for developing appropriate conservation and management strategies. We previously established an age estimation method for brown bears 43 and other bear species 44 based on DNA methylation levels in the blood; however, this approach requires capture and anesthesia, which are inherently invasive. To address this limitation, a less-invasive technique using hair DNA was recently developed 45 . In this study, we aimed to establish a non-invasive and accurate age estimation method for brown bears based on fecal DNA methylation levels. Materials and Methods Ethical Statement All procedures involved in sample collection from live animals in the Shiretoko National Park (incl. a Special Wildlife Protection Area) were conducted in accordance with the Guidelines for Animal Care and Use, Hokkaido University, and were approved by the Animal Care and Use Committee of the Graduate School of Veterinary Medicine, Hokkaido University (permit nos: 1106, 1151, 1152, 15009, 17005, 18-0083, 19-0021, 20-0146, and 23-0014) and by the Hokkaido Regional Environment Office and Kushiro Nature Conservation Office of the Ministry of the Environment, Japan (Permit nos: 2105071, 2205061, 2305161, 2405291, and 2504221). In addition, all methods were carried out in compliance with ARRIVE guidelines. Study area and sampling Wild bears The present study was conducted on both wild and captive brown bears. Wild bears were sampled from the Shiretoko Peninsula in eastern Hokkaido, Japan (Supplementary Figure S_M1; 43°50′–44°20′ N, 144°45′–145°20′ E). An area extending from the middle of the peninsula to its tip, covering 610 km², has been designated as a national park. An area covering 711 km², including the national park and the surrounding terrestrial and marine zones, was designated as a UNESCO World Natural Heritage Site in July 2005. Fecal samples collected throughout the peninsula were roughly divided into two groups according to sampling site. The first sample was collected from the Rusha area of the national park (Supplementary Figure S_M1; 44°11′–44°12′ N, 145°10′–145°12′ E). This area consists of a narrow estuarine coastline approximately 3 km long, and is designated as a Special Wildlife Protection Area, where public access is prohibited without permission. There were no permanent residents except for one fisherman’s settlement. As fishermen have not excluded bears from this area for several decades, they have become habituated to the presence of humans, allowing direct observations at close range. Since the late 1990s, monitoring surveys of identifiable bears based on their external appearance have been conducted and since 2009, DNA samples have been collected from these individuals 34,46 . The second set of fecal samples was collected from other areas across the peninsula, including Shari and Rausu towns (Supplementary Figure S_M1). In these areas, management surveys of brown bears have been conducted in response to human-bear conflicts near residential and tourist zones. For some individuals, the actual age was identified through visual observation and individual identification based on DNA analyses 46,47 . Fecal samples collected from these two areas between 2021 and 2025 were used in this study. Captive bears Fecal samples were collected from captive brown bears in July 2025 at Noboribetsu Bear Park (Noboribetsu, Hokkaido, Japan) and the Sapporo Maruyama Zoo (Sapporo, Hokkaido, Japan) (Supplementary Figure S_M1). At the Noboribetsu Bear Park, fecal samples were collected from 31 bears (13 males and 18 females) ranging in age from 0 to 34 years. Among them, two individuals were rescued from the wild as cubs and have since been kept at the facility, whereas all other individuals were born in captivity. When multiple individuals were kept in the same enclosure, multiple feces found within the enclosure were collected. In addition, to ensure sufficient sample numbers, multiple samples were collected from a single fecal deposit. At the Sapporo Maruyama Zoo, two brown bears (a 14-year-old male and a 17-year-old female) were kept, both born in captivity. In the facility, each bear was housed in a separate enclosure. At both facilities, fecal sampling was conducted during routine daily cleaning, ensuring that all collected feces were less than one day old after defecation. Sample collection Fecal samples were collected according to previously reported methods 47 . For wild bears, only feces estimated to have been defecated within five days were selected, although accurately determining freshness was almost impossible in most cases. The entire surface of each fecal sample was carefully scrubbed using a disposable flocked swab (Flocked Swab R30, Sugiyama-gen Co., ltd., Tokyo, Japan), which was then immersed in a 2.0 mL tube containing 1.1 mL of InhibitEX Buffer (Qiagen inc., Hilden, Germany). This process was repeated several times for each sample. Samples were stored at −20℃ until DNA extraction. DNA extraction and individual identification DNA was extracted using the QIAamp DNA Stool Mini Kit (Qiagen inc.) according to the manufacturer’s protocol and eluted in 50 µL of elution buffer. To verify the success of DNA extraction and identify individuals, microsatellite analysis was conducted for all fecal DNA samples (i.e., both wild and captive). Based on our previous studies 34,47 , two primer mixtures were used, each containing three microsatellite loci (Primer mix A: G1A, MU05, and MU51; Primer mix B: MU50, G10B, and MU23) and a sex identification marker (Amelogenin) 48 . Polymerase chain reaction (PCR) amplification was performed using a Multiplex PCR Assay Kit (Takara Bio Inc., Shiga, Japan) in a total reaction volume of 10 µL containing 1 µL of genomic DNA, 0.05 µL of Kit Mix 1, 5 µL of Kit Mix 2, 0.5 µL of primer mix (0.25 µmol/L each), and 3.45 µL of PCR-grade water. The PCR conditions were as follows: 95℃ for 30 sec, followed by 40 cycles of 95℃ for 30 sec, 55℃ for 30 sec, and 72℃ for 30 sec, with a final extension at 72℃ for 5 min. The amplified products were analyzed using a SeqStudio Genetic Analyzer (Thermo Fisher Scientific, Waltham, MA, USA) to determine genotypes. As for samples collected in the wild, each identified genotype was compared with previously genotyped individuals from the Shiretoko population to confirm their identity and chronological age, and samples from a subset of these individuals were used. Individual identification was performed for samples collected from enclosures in which multiple bears were housed in captivity. DNA purification Among the samples with confirmed chronological ages, those that showed stable PCR amplification in microsatellite analysis were purified using the FastGene Gel/PCR Extraction Kit (Nippon Genetics Co., Ltd., Tokyo, Japan). Even if individual identification was possible, samples showing unstable PCR amplification were excluded from subsequent analyses. The purified DNA was eluted in 30–40 µL of elution buffer. In some cases, two samples derived from the same individual collected on the same day were combined before purification to increase the amount of DNA. Subsequently, DNA concentration was measured using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific). Bisulfite conversion Bisulfite treatment was performed using the EZ DNA Methylation-Gold Kit (Zymo Research, Irvine, CA, USA), according to the manufacturer’s protocol. In most cases, 20 µL of DNA solution was added. For samples in which the DNA amount per 20 µL exceeded the maximum processing amount (2,000 ng), the sample was diluted with molecular-grade water to adjust the added DNA amount to 2,000 ng. The volume of elution buffer was adjusted according to the amount of processed DNA to achieve a final DNA concentration of 10 ng/µL. Selection of target genomic locations Two genomic locations adjacent to DLX5 (Distal-less homeobox 5) and SLC12A5 (Solute carrier family 12 member 5) genes (hereafter referred to as DLX5 and SLC12A5 , respectively) were selected as targets for methylation level analysis. The DLX5 is involved in bone development 49 , and CpG sites adjacent to this gene have been reported to correlate with age in multiple felid species 30 . The SLC12A5 encodes the neuron-specific K⁺/Cl⁻ cotransporter KCC2, which mediates rapid hyperpolarization through GABA signaling 50 . The CpG sites adjacent to SLC12A5 in the blood of multiple bear species have been reported to correlate with age 43,44 . For DLX5 , the homologous sequences containing the target CpG sites in the brown bear genome were identified using the Basic Local Alignment Search Tool (BLAST) provided by the National Center for Biotechnology Information (NCBI). For SLC12A5 , the CpG sites reported by Nakamura et al. 43 were used. Detailed information on each genomic region is presented in Table 1. Polymerase chain reaction and pyrosequencing PCR and subsequent pyrosequencing were conducted as described by Yamazaki et al. 51 . A sample list for the methylation level analysis is shown in Supplementary Table S_M1. PCR was performed using the TaKaRa EpiTaq HS (for bisulfite-treated DNA; Takara Bio Inc.). PCR and pyrosequencing primers were designed using Methyl Primer Express v1.0 (Thermo Fisher Scientific) and PyroMark Assay Design v2.0.2.5 (Qiagen inc.). Details of each primer are listed in Supplementary Table S_M2. PCR was performed in two steps such that biotin-modified primers could be applied to both target genomic locations 51 . The first PCR was performed in a total reaction volume of 15 µL containing 1 µL of genomic DNA (diluted to contain 10 ng DNA), 0.075 µL TaKaRa EpiTaq HS, 1.5 µL 10 × EpiTaq PCR Buffer (Mg²⁺-free), 1.5 µL MgCl₂, 1.8 µL dNTP mixture, 0.3 µL each of the forward and reverse primers (10 µmol/L), and 8.525 µL molecular-grade water. The PCR conditions were 98℃ for 1 min, followed by 40 cycles of 98℃ for 10 sec, the annealing temperature (listed in Supplementary Table S_M2) for 30 sec, and 72℃ for 30 sec. Each sample was analyzed in triplicate. The second PCR was performed in a total reaction volume of 15.13 µL containing 0.1 µL of the first PCR product, 0.075 µL TaKaRa EpiTaq HS, 1.5 µL 10 × EpiTaq PCR Buffer (Mg²⁺- free), 1.5 µL MgCl₂, 1.8 µL dNTP mixture, 0.3 µL of the forward primer (10 µmol/L), 0.06 µL of the reverse primer (10 µmol/L), 0.27 µL of the biotin-modified primer (10 µmol/L), and 9.525 µL molecular-grade water. The PCR conditions were 98℃ for 1 min, followed by 35 cycles of 98℃ for 10 sec, the annealing temperature (listed in Supplementary Table S_M2) for 30 sec, and 72℃ for 30 sec. The first and second PCR products were electrophoresed on a 2% agarose gel to confirm the successful amplification of the target region. The methylation levels of the target CpG sites (3 and 4 CpGs for DLX1 and SLC12A5 , respectively; Supplementary Figure S_M2) were determined using the PyroMark Q48 software with PyroMark Q48 Advanced Reagents (Qiagen inc.), according to the manufacturer’s protocol. Samples showing unstable baselines in the pyrogram were recommended for reanalysis because the methylation levels could not be correctly calculated 52 . However, because fecal DNA is often of low quality and quantity and to maintain procedural simplicity, these samples were not re-analyzed and were treated as failed analyses. Finally, samples were considered successfully analyzed when at least two of the three replicate measurements satisfied the following two criteria 53,54 : (1) The amplification products of the first and second PCR were confirmed by electrophoresis. 2) In pyrosequencing, the pyrogram showed a stable baseline. For samples that met these criteria, the average methylation level from two or three successful replicates was used as the final methylation level. Determination of host DNA copy number using quantitative PCR Because fecal DNA contains not only host DNA, but also intestinal bacteria and dietary DNA, the measured DNA concentration does not necessarily represent the host DNA concentration. Therefore, to quantify the number of host DNA copies used for methylation level analysis, quantitative PCR (qPCR) was performed. The amplification target, the intron 12 region of the coagulation factor II ( F2 ) gene was selected according to a previous study in which host DNA was quantified using qPCR was developed in polar bears ( Ursus maritimus ) 55 . The F2 gene encodes a protein involved in fibrin clot formation 56 , and owing to sequence variations within the intron region, enables species-specific detection 57 . Using BLAST provided by NCBI, we confirmed that the sequence of the target region in polar bears (NW_007907185.1; positions 2,099,674–2,099,774) was identical to the homologous sequence in brown bears (NW_020656153.1; positions 1,273,581–1,273,681). Primers with the same sequences described by Hayward et al. 55 (Supplementary Table S_M2) were used. A plasmid containing a 101 bp sequence including the primer attachment region was synthesized (FASMAC, Kanagawa, Japan), and serially diluted to 10⁷, 10⁶, 10⁵, 10⁴, 10³, 10², and 10¹ copies to construct a standard curve. Details of the synthetic plasmids are listed in Supplementary Table S_M3. Standard curves and DNA copy numbers were analyzed using a StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). qPCR was performed using THUNDERBIRD™ SYBR® qPCR Mix (TOYOBO Co., Ltd., Osaka, Japan). Genomic DNA extracted from the blood of two brown bears was included in all qPCR runs as a positive control, and a negative control was included in each run. The PCR was performed in a total reaction volume of 10 µL containing 1 µL of genomic DNA prior to bisulfite conversion, 5 µL THUNDERBIRD SYBR qPCR Mix, 0.2 µL 50×ROX reference dye, 0.2 µL each of the forward and reverse primers (10 µmol/L), and 3.4 µL molecular-grade water. The PCR conditions were 95℃ for 1 min, followed by 40 cycles of 95℃ for 15 sec, 60℃ for 30 sec, and 72℃ for 30 sec, followed by a melting curve analysis (95℃ for 15 sec, 60℃ for 1 min, and 95℃ for 15 sec). Each sample was analyzed in duplicate. The average values of the duplicates were calculated to determine the number of host DNA copies used for the analysis (i.e., the host DNA copy number added to the first PCR). Age estimation model and model validation To construct a robust age estimation model, only samples meeting the following conditions were used: 1) methylation levels in both the DLX5 and SLC12A5 regions were successfully analyzed, and 2) the number of host DNA copies used for analysis was at least 1,392 or higher. This threshold was based on a previous study 58 reporting that in human methylation analyses, a sample containing ≥ 5 ng host DNA (approximately 1,392 human DNA molecules) is sufficient to obtain stable methylation measurements. If multiple samples from the same individual met these criteria, the sample with the highest number of host DNA copies used for analysis was selected for model construction. In other words, only one sample per individual was used for model construction and the remaining samples were used for model performance evaluation, as described later. Based on the pyrosequencing results, three types of age estimation models were generated following previous studies 43-45 : a single regression model requiring the methylation level of a single CpG site and two multiple regression models (elastic net regression and support vector regression [SVR]) that utilize multiple CpG methylation levels. Age and DNA methylation levels used for model construction were standardized to a mean of 0 and a standard deviation of 1 before integration into the models. Single regression models were generated using the R command “lm.” Elastic net regression models, a type of penalized regression methods frequently used for age estimation in other species 12 , were constructed using the R package “glmnet.” Optimal parameters (alpha and lambda) were determined using “cv.glmnet”. The SVR models, which have been suggested to outperform elastic net regression for age estimation 59,60 , were constructed using the R package “e1071.” The parameters (cost, gamma, and epsilon) were optimized with the “tune” command under the fixed settings “type = eps-regression, kernel = radial”. CpG site selection for the SVR models was based on the absolute values of regression coefficients (β) obtained from the elastic net regression and was performed by sequentially excluding CpG sites with lower contribution. Additionally, for practical applications, the SVR models using a single region ( DLX5 or SLC12A5 ) were also constructed. All models were validated using leave-one-out cross-validation (LOOCV), a method that extracts one sample as a test set, while using the remaining samples for training, which is repeated as many times as the number of samples. Model accuracy was evaluated by calculating the mean absolute error (MAE), median absolute error (MedAE), and root mean square error (RMSE) based on the LOOCV results, and the three models were compared. Model performance evaluation To evaluate the performance of the constructed models, age estimation was performed using samples that were not included in model construction. These samples included those that met the PCR amplification and pyrogram criteria but had fewer than 1,392 host DNA copies used for analysis, as well as samples not used for model construction because of overlap among individuals. The best-performing model was then used to estimate the ages of the samples not included in model construction, and the estimated ages were compared with the chronological ages. Generalized linear mixed models (GLMMs) were used to assess whether chronological age, number of host DNA copies used for analysis, and growth environment (i.e., captive or wild) affected the estimation error. Some samples that were not included in the model construction were collected from the same individual. Considering the dependence between samples, individual names were used as random effects. Δage (estimated age − chronological age) and |Δage| (absolute difference between estimated and chronological age) were used as dependent variables, and the three factors (chronological age, the number of host DNA copies used for analysis, and growth environment) and their interactions were included as explanatory variables. Sex was not included as an explanatory variable because older individuals used for model construction were biased toward females (i.e., among individuals aged ≥15 years, five were male and ten were female). To evaluate whether the host DNA copy number affected the success of methylation level analysis (i.e., fulfilling the aforementioned criteria in PCR and pyrosequencing), the number of host DNA copies used for analysis was statistically compared between successful and failed cases using the Wilcoxon rank-sum test. The significance level was set at 5%. This comparison included all fecal DNA samples used for the methylation analysis, regardless of whether they were incorporated into the model construction. Results Sample selection The samples were selected using the process shown in Figure 1. In total, 25 fecal DNA samples from wild bears (aged 0–28 years) and 18 samples from captive bears (aged 0–26 years) were included in model construction. Additionally, 18 samples from wild bears (10 individuals, aged 1–20 years) and 10 samples from captive bears (8 individuals, aged 1–19 years) were successfully analyzed for DNA methylation levels and were included in the model performance evaluation (i.e., not included in model construction), due to multiple samples from the same individual or insufficient host DNA copy numbers (≤1,392). Detailed information regarding each step is provided in Supplementary Table S_R1. Correlation between DNA methylation level and chronological age The relationships between DNA methylation levels and the chronological age of the 43 samples included in the model construction are shown in Figure 2. Significant positive correlations were observed for all the CpG sites (Supplementary Figure S_M2; DLX5-1: correlation coefficient [cor] = 0.894, p < 0.001; DLX5-2: cor = 0.909, p < 0.001; DLX5-3: cor = 0.850, p < 0.001; SLC12A5-1: cor = 0.895, p < 0.001; SLC12A5-2: cor = 0.911, p < 0.001; SLC12A5-3: cor = 0.917, p < 0.001; SLC12A5-4: cor = 0.922, p < 0.001). Among these, SLC12A5-4 showed the strongest correlation, although all CpG sites exhibited cor ≥ 0.85. Age estimation model Comparison of models Based on the methylation levels at the seven CpG sites, three age estimation models were constructed. The single regression model using the methylation level of SLC12A5-3 showed the highest accuracy, with an MAE of 2.33 years after LOOCV. The formula for age estimation is as follows: “Estimated Age” = (−2.133e−11 + 0.9171 × “methylation level of SLC12A5-3”) × 8.274 (standard deviation of training data) + 10.086 (mean of training data). The elastic net regression model in which DLX5-3 was excluded because its regression coefficient was zero exhibited the best performance when DLX5-1, DLX5-2, SLC12A5-1, SLC12A5-2, SLC12A5-3, and SLC12A5-4 were used. The MAE after LOOCV is 2.35 years. The formula for age estimation is as follows: “Estimated Age” = (2.32e−11 + 0.1033 × “methylation level of DLX5-1” + 0.1577 × “methylation level of DLX5-2” + 0.0990 × “methylation level of SLC12A5-1” + 0.1414 × “methylation level of SLC12A5-2” + 0.1628 × “methylation level of SLC12A5-3” + 0.2348 × “methylation level of SLC12A5-4”) × 8.274 + 10.086. The support vector regression (SVR) models using DLX5-2, SLC12A5-2, SLC12A5-3, and SLC12A5-4 exhibited the highest accuracy, with an MAE of 2.08 years after LOOCV. The detailed statistics of each model are presented in Table 2, and the relationships between the chronological and estimated ages after LOOCV are shown in Figure 3. The regression coefficients of the elastic net model and the parameter settings for the elastic net regression and SVR models are summarized in Supplementary Tables S_R2 and S_R3. The R script used for age estimation is provided in the Supplementary Materials. Model performance evaluation The SVR model using DLX5-2 and SLC12A5-2, -3, and -4, which exhibited the lowest MAE after LOOCV, was selected as the best model. When applied to samples that were not included in the model construction, the model achieved an MAE of 2.10, a MedAE of 1.35, and an RMSE of 2.93. The relationship between chronological and estimated ages and the results of the generalized linear mixed models (GLMMs) are shown in Supplementary Figure S_R1 and Table 3–4, respectively. Neither chronological age nor growth environment had a significant effect on either Δage (predicted age − chronological age) or |Δage| (absolute difference between predicted and chronological age). Samples with lower host copy numbers used for analysis tended to show larger prediction errors (i.e., |Δage|), although they were not statistically significant (p = 0.09). None of the interactions among the explanatory variables were statistically significant. In addition, samples in which the methylation level analysis failed had significantly lower host DNA copy numbers used for analysis than successful samples (p < 0.001; Figure 4). Discussion This is the first study to establish a non-invasive method for epigenetic age estimation using fecal DNA from bears. Based on the MAE after LOOCV, it was suggested that the best model was the SVR model constructed using four CpG sites (i.e., DLX5-2 and SLC12A5-2, -3, and -4) with an MAE of 2.08 years (6.9–10.4% of lifespan). This accuracy was slightly inferior to the age estimation models developed using blood DNA of brown bears (MAE: 1.30; 4.3–6.5% of lifespan 43 ), but superior to the hair DNA-based model (MAE: 3.19; 10.6–16.0% of lifespan 45 ). In previous studies using fecal DNA, the reported MAE values were 5.08 years for Indo-Pacific bottlenose dolphins (lifespan: approx. 40–50 years; about 10.2–12.7% of lifespan 28 ), 2.54 years for Tsushima leopard cats (lifespan: 15–20 years; 12.7–16.9% of lifespan 30 ), 26 days for mice (lifespan: 30–32 months; 3% of lifespan 29 ). Compared to these species, the current model achieved comparable or higher accuracy for brown bears (lifespan: 20–30 years 33 ). When comparing the SVR models constructed using only SLC12A5 (SLC12A5-2, -3, and -4) to the best model, the difference in MAE was negligible (2.11 vs. 2.08 years) and MedAE values were slightly superior in the former case (0.92 vs. 0.99 years). Therefore, when analyzing many samples, a model constructed using three CpG sites in SLC12A5 , which requires the amplification of only one genomic region, is more practical and cost-effective. The CpG sites adjacent to SLC12A5 have also been used as a single-target locus in an age estimation model for blood DNA 43,44 and as the best-performing target in a fecal DNA-based model for Tsushima leopard cats 30 . These results suggest that this genomic location may serve as a valuable target for developing DNA methylation-based age estimation models across different sample types and mammalian species. Age estimation using fecal DNA has several challenges, including the presence of PCR inhibitors in feces that can hinder DNA amplification 61,62 , unavoidable contamination by DNA derived from diet and intestinal bacteria, and degradation of host DNA quantity and quality due to environmental exposure 31,32 . In this study, we purified all samples to minimize the inclusion of inhibitory components and remove short, fragmented DNA. Furthermore, quantification of the host DNA copy numbers used for the analysis clarified their impact on the success or failure of methylation-level measurements. These newly added steps were not present in previous fecal DNA-based studies 28-30 and significantly contributed to the success of the analysis and improvement of its accuracy. The number of host DNA copies used for analysis is a clear factor that determines the success or failure of the methylation-level analysis. In addition, there was a tendency that the estimation error (i.e., |Δage|) to increase as the host DNA copy number decreased. This suggests that for biological samples, such as fecal DNA, which often present challenges in DNA quantity and quality, pre-quantification of host DNA copy number may enable prior decisions on sample suitability for analysis, thereby enhancing cost-effectiveness in practical applications. In the current study’s age estimation model construction, a threshold of 1,392 host DNA copies used for analysis was set to ensure model robustness 58 . Nevertheless, among the 30 samples that did not meet this criterion, methylation level analysis was successful in 10 samples (the lowest host DNA copy number used for analysis among the successful samples was 172). This indicates that even samples below the threshold can be used for methylation level estimation if it is possible to include approximately 150 host DNA copies or more in the first PCR and successful amplification is confirmed by electrophoresis. However, it should be kept in mind that there is an increased risk of estimation error with a lower host DNA copy number. Intron 12 of the F2 gene, targeted for qPCR quantification of host DNA, has been shown to be unaffected by DNA derived from prey species of polar bears, including the American moose ( Alces alces ), Arctic hares ( Lepus arcticus ), spotted seals ( Phoca largha ), ringed seals ( Pusa hispida ), and Arctic foxes ( Vulpes lagopus ), and is not affected by contamination with human DNA 55 . Nonetheless, because cannibalism, including infanticide, has been documented in brown bears 63 , the present method may not be applicable to feces containing bear carcasses. The use of fecal DNA for age estimation is highly applicable. One major advantage is the potential to improve the accuracy of population age structure and population dynamic assessments. Population monitoring of bears using fecal samples has been conducted worldwide 64,65 . Similar studies have been conducted on other wildlife species, including wolves 66 , elephants ( Loxodonta Africana ) 67 , snow leopards ( Panthera uncia ) 68 , and white-tailed deer ( Odocoileus virginianus ) 69 . These studies used DNA information obtained from samples to infer population dynamics using capture-mark-recapture methods or spatial capture-recapture approaches. However, DNA information alone does not provide the age structure of a population, requiring additional methods, such as camera trapping or direct visual surveys 70,71 . By applying the method established in this study, the age structure of a population can be inferred from fecal samples. It is essential to identify age classes such as juvenile and reproductively mature individuals to make more accurate predictions of future population trends 72,73 . This study provides valuable information for population management and conservation. Feces also serve as biological samples that provide a wide range of valuable information for understanding the life histories of animals. A representative example of brown bears is dietary information 39,74 . These studies analyzed diet using DNA barcoding and identified food remains in fecal samples. Although such approaches can reveal regional and seasonal differences in diet, age-related dietary changes have relied on indirect methods such as stable isotope analyses using biological samples (e.g., hair 75 , teeth 76 , and blood 77 ). By combining fecal content analysis with age estimation, it is possible to clarify age-related shifts in resource use and foraging strategies, allowing for a deeper understanding of the life history of brown bears. In addition, determining age-specific dependence on resources can help identify key environments that must be conserved to maintain population size, thereby making a substantial contribution to wildlife conservation 78 . Similarly, other applications are expected, such as studies on the gut microbiome 79 , analyses of fecal hormones (e.g., fecal estradiol and progesterone 80 and fecal glucocorticoids 81,82 ), research on gastrointestinal parasites 83 , and infectious diseases 84 . The methodology developed in this study elucidates age-dependent biological changes, including nutrition, reproduction, stress, disease infection, and immunity. This study provides valuable insights into the adaptive strategies and physiological changes in wild populations. One of the limitations of this study was that sex differences could not be fully evaluated because of sampling bias. In the samples used for model construction, older individuals were biased toward females (two-thirds of the individuals aged ≥15 years were females). The shortage of samples from adult males, particularly wild males, was notable; only five males were included in the model construction, and all were 4 years old or younger. Wild male bears do not remain in their natal areas and typically disperse at approximately three years of age, with dispersal distances longer than those of females 85,86 . Therefore, it was not possible to collect fecal samples from adult males born within the study area whose ages were known. Another challenge that needs to be addressed is determining whether the present model can be applied to other brown bear populations. The samples used for the model construction were limited to wild individuals from a restricted area in Hokkaido and two captive facilities, all of which were one subspecies of the brown bear. Brown bears are among the most widely distributed terrestrial mammalian species, occupying broad regions across North America, Europe, and Asia. They inhabit diverse environments, including forests, coastal zones, savannas, desert areas, and alpine zones, and exhibit substantial regional variations in diet and body size 87,88 . For example, a study of three North American populations showed that, even within North America, coastal and inland groups differ in the principal food items that influence body size 89 . Although this study found no effect of environment (i.e., captive or wild) on the estimation error, it remains necessary to verify whether the current age estimation model can be applied to other brown bear populations without calibration. In conclusion, this study established the first epigenetic clock for brown bears using fecal DNA samples based on the methylation levels of CpG sites adjacent to the DLX5 and SLC12A5 genes. The best model achieved an MAE of 2.08 years, demonstrating high accuracy. Fecal samples can be collected non-invasively, which is a clear distinction from existing techniques for estimating the age of brown bears and other species based on blood, hair, and other sample types. Furthermore, a new process was developed that enables the preliminary selection and precise evaluation of samples by quantifying the number of host DNA copies present in fecal DNA. This approach addresses issues specific to fecal DNA, such as contamination by nonhost DNA and degradation of DNA quality due to environmental exposure, which are essential considerations for practical applications in field research. This study holds substantial value for ecological research and conservation of brown bears and is expected to contribute to the development of non-invasive approaches for other species. Declarations Acknowledgements We would like to express our profound gratitude and appreciation to all the people who cooperated with the sampling, especially to all the staff of the Shiretoko Nature Foundation, Noboribetsu bear park, and Sapporo Maruyama Zoo (especially to Mr. Wataru Goshima and Mr. Michiaki Shimizu) for providing bear fecal samples. We wish to thank Hatsusaburo Ose and all the members of the Shiretoko Fishery Productive Association for their kind support. Finally, we thank Editage (www.editage.jp) for English language editing. Funding This study was supported by funding from the Japan Society for the Promotion of Science (JSPS) (https://www.jsps.go.jp/english/e-grants/index.html) KAKENHI (grant numbers: JP19K06833, JP22K14910, JP23K05312, JP24KJ0304, JP25K22873 and JP25H01002), the Collaborative Research Program of Wildlife Research Center, Kyoto University (grant no. 2021-A-16), and Grant for Basic Science Research Projects from The Sumitomo Foundation (grant no. 200561). This research was performed by the Environment Research and Technology Development Fund (JPMEERF20254002) of the Environmental Restoration and Conservation Agency provided by Ministry of the Environment of Japan. Author Contributions S.O., S.N., J.Y. and M.S. designed the study. S.O., S.N., and M.S. performed laboratory work, and constructed each age estimation model. S.O. S.N., K.H., Y.Y., N.M., M.Y., M.N., S.W., and M.S. were involved in sample collection. J.Y. and H.I., and M. I-M. supported the technical aspects of the experiment. S.O. and M.S. wrote the article with inputs from S.N., J.Y., H.I., and T.T. All authors reviewed the article. Data availability The datasets and R-scripts used in this study are available from the corresponding authors on reasonable request. Competing interests The authors declare no competing interests. References Colchero, F. et al. The diversity of population responses to environmental change. Ecol. Lett. 22 , 342–353. https://doi.org/10.1111/ele.13195 (2019). Robert, A., Chantepie, S., Pavard, S., Sarrazin, F. & Teplitsky, C. Actuarial senescence can increase the risk of extinction of mammal populations. Ecol. Appl. 25 , 116–124. 10.1890/14-0221.1 (2015). Tidière, M. et al. Comparative analyses of longevity and senescence reveal variable survival benefits of living in zoos across mammals. Sci. Rep. 6 , 36361. 10.1038/srep36361 (2016). Thomas, D. C. Metachromatic staining of dental cementum for mammalian age-determination. J. Wildl. Manage. 41 , 207–210 (1977). doi:Doi 10.2307/3800596. McLaughlin, C. R. et al. Precision and accuracy of estimating age of maine black bears by cementum annuli. Int. Conf. Bear. Res. Manage. 8 , 415–419. 10.2307/3872945 (1990). Rolandsen, C. M. et al. Accuracy and repeatability of moose ( Alces alces ) age as estimated from dental cement layers. Eur. J. Wildl. Res. 54 , 6–14. 10.1007/s10344-007-0100-8 (2008). Bocklandt, S. et al. Epigenetic predictor of age. Plos One . 6 , e14821. 10.1371/journal.pone.0014821 (2011). Bogdanovic, O. & Veenstra, G. J. C. DNA methylation and methyl-CpG binding proteins: developmental requirements and function. Chromosoma 118 , 549–565. 10.1007/s00412-009-0221-9 (2009). Moore, L. D., Le, T. & Fan, G. P. DNA methylation and its basic function. Neuropsychopharmacol 38 , 23–38. 10.1038/npp.2012.112 (2013). Jones, M. J., Goodman, S. J. & Kobor, M. S. DNA methylation and healthy human aging. Aging Cell. 14 , 924–932. 10.1111/acel.12349 (2015). Lu, A. T. et al. Universal DNA methylation age across mammalian tissues. Nat. Aging . 3 , 1462–1462. 10.1038/s43587-023-00499-7 (2023). Horvath, S. DNA methylation age of human tissues and cell types. Genome Biol. 14 , R115. 10.1186/gb-2013-14-10-r115 (2013). Bekaert, B., Kamalandua, A., Zapico, S. C., Van de Voorde, W. & Decorte, R. Improved age determination of blood and teeth samples using a selected set of DNA methylation markers. Epigenetics 10 , 922–930. 10.1080/15592294.2015.1080413 (2015). Voisin, S. et al. An epigenetic clock for human skeletal muscle. J. Cachexia Sarcopenia Muscle . 11 , 887–898. 10.1002/jcsm.12556 (2020). Hong, S. R. et al. DNA methylation-based age prediction from saliva: High age predictability by combination of 7 CpG markers. Forensic Sci. International: Genet. 29 , 118–125. 10.1016/j.fsigen.2017.04.006 (2017). Stubbs, T. M. et al. Multi-tissue DNA methylation age predictor in mouse. Genome Biol. 18 , 68. 10.1186/s13059-017-1203-5 (2017). Lowe, R. et al. DNA methylation clocks as a predictor for ageing and age estimation in naked mole-rats,. Aging 12, 4394–4406, (2020). 10.18632/aging.102892 Thompson, M. J., Vonholdt, B., Horvath, S. & Pellegrini, M. An epigenetic aging clock for dogs and wolves. Aging 9 , 1055–1068. 10.18632/aging.101211 (2017). Qi, H. et al. Age estimation using methylation-sensitive high-resolution melting (MS-HRM) in both healthy felines and those with chronic kidney disease. Sci. Rep. 11 , 19963. 10.1038/s41598-021-99424-4 (2021). Raj, K. et al. Epigenetic clock and methylation studies in cats. Geroscience 43 , 2363–2378. 10.1007/s11357-021-00445-8 (2021). Caulton, A. et al. Development of epigenetic clocks for key ruminant species. Genes 13 10.3390/genes13010096 (2021). Polanowski, A. M., Robbins, J., Chandler, D. & Jarman, S. N. Epigenetic estimation of age in humpback whales. Mol. Ecol. Resour. 14 , 976–987. 10.1111/1755-0998.12247 (2014). Beal, A. P., Kiszka, J. J., Wells, R. S. & Eirin-Lopez, J. M. The bottlenose dolphin epigenetic aging tool (BEAT): A molecular age estimation tool for small cetaceans. Front. Mar. Sci. 6 , 561. 10.3389/fmars.2019.00561 (2019). Ito, H., Udono, T., Hirata, S. & Inoue-Murayama, M. Estimation of chimpanzee age based on DNA methylation. Sci. Rep. 8 , 9998. 10.1038/s41598-018-28318-9 (2018). Lemaitre, J. F. et al. DNA methylation as a tool to explore ageing in wild roe deer populations. Mol. Ecol. Resour. 22 , 1002–1015. 10.1111/1755-0998.13533 (2022). Petkovich, D. A. et al. Using DNA methylation profiling to evaluate biological age and longevity interventions. Cell. Metab. 25 , 954–960e956. 10.1016/j.cmet.2017.03.016 (2017). Marrero, P., Fregel, R., Cabrera, V. M. & Nogales, M. Extraction of high-quality host DNA from feces and regurgitated seeds: a useful tool for vertebrate ecological studies. Biol. Res. 42 , 147–151 (2009). Yagi, G. et al. Non-invasive age estimation based on faecal DNA using methylation-sensitive high-resolution melting for Indo-Pacific bottlenose dolphins. Mol. Ecol. Resour. 24 , e13906. 10.1111/1755-0998.13906 (2024). Hanski, E. et al. Epigenetic age estimation of wild mice using faecal samples. Mol. Ecol. 33 , e17330. 10.1111/mec.17330 (2024). Qi, H. Development of epigenetic clocks in multiple felid species — from small to big, domestic to wild. Doctral thesis, Kyoto University, (2023). 10.14989/doctor.k25151 Agetsuma-Yanagihara, Y., Inoue, E. & Agetsuma, N. Effects of time and environmental conditions on the quality of DNA extracted from fecal samples for genotyping of wild deer in a warm temperate broad-leaved forest. Mammal Res. 62 , 201–207. 10.1007/s13364-016-0305-x (2017). Brinkman, T. J., Schwartz, M. K., Person, D. K., Pilgrim, K. L. & Hundertmark, K. J. Effects of time and rainfall on PCR success using DNA extracted from deer fecal pellets. Conserv. Genet. 11 , 1547–1552. 10.1007/s10592-009-9928-7 (2009). Bunnell, F. L. & Trait, D. E. N. in Dynamics of Large Mammal Populations (eds C. W. Fowler & T. D. Smith) Ch. 4, 75–98 (Wiley-Interscience, (1981). Shimozuru, M. et al. Reproductive parameters and cub survival of brown bears in the Rusha area of the Shiretoko Peninsula, Hokkaido, Japan. Plos One . 12 , e0176251. 10.1371/journal.pone.0176251 (2017). Shirane, Y. et al. Development of a noninvasive photograph-based method for the evaluation of body condition in free-ranging brown bears. Peerj 8 , e9982. 10.7717/peerj.9982 (2020). Mano, T. & Tsubota, T. Reproductive characteristics of brown bears on the Oshima Peninsula, Hokkaido, Japan. J. Mammal . 83 , 1026–1034. 10.1644/1545-1542(2002)083%3C1026 (2002). :Rcobbo>2.0.Co;2 González-Bernardo, E., Russo, L. F., Valderrábano, E., Fernández, A. & Penteriani, V. Denning in brown bears. Ecol. Evol. 10 , 6844–6862. 10.1002/ece3.6372 (2020). Friebe, A. et al. Factors affecting date of implantation, parturition, and den entry estimated from activity and body temperature in free-ranging brown bears. Plos One . 9 , e101410. 10.1371/journal.pone.0101410 (2014). Shirane, Y. et al. Dining from the coast to the summit: Salmon and pine nuts determine the summer body condition of female brown bears on the Shiretoko Peninsula. Ecol. Evol. 11 , 5204–5219. 10.1002/ece3.7410 (2021). Miller, S. D. Population management of bears in North America. Bears: Their Biology Manage. 8 , 357–373. 10.2307/3872940 (1990). Bombieri, G. et al. Brown bear attacks on humans: a worldwide perspective. Sci. Rep. 9 , 8573. 10.1038/s41598-019-44341-w (2019). Can, Ö. E., D'Cruze, N., Garshelis, D. L., Beecham, J. & Macdonald, D. W. Resolving human-bear conflict: A global survey of countries, experts, and key factors. Conserv. Lett. 7 , 501–513. 10.1111/conl.12117 (2014). Nakamura, S. et al. Age estimation based on blood DNA methylation levels in brown bears. Mol. Ecol. Resour. 23 , 1211–1225. 10.1111/1755-0998.13788 (2023). Shimozuru, M. et al. Epigenetic clock in bears: A simple cost-effective blood DNA methylation-based age estimation method applicable to multiple bear species. Ecol. Evol. 15 , e71424. 10.1002/ece3.71424 (2025). Nakamura, S. et al. Less-invasive age estimation using hair based on DNA methylation in brown bears. Sci. Rep-Uk . 15 , 45696. 10.1038/s41598-025-27455-2 (2025). Shimozuru, M. et al. Incidence of multiple paternity and inbreeding in high-density brown bear populations on the Shiretoko Peninsula, Hokkaido, Japan. J. Hered . 110 , 321–331. 10.1093/jhered/esz002 (2019). Shimozuru, M. et al. Estimation of breeding population size using DNA-based pedigree reconstruction in brown bears. Ecol. Evol. 12 , e9246. 10.1002/ece3.9246 (2022). Yamamoto, K. et al. Sex identification of Japanese black bear, Ursus thibetanus japonicus , by PCR based on amelogenin gene. J. Vet. Med. Sci. 64 , 505–508. 10.1292/jvms.64.505 (2002). Heo, J. S., Lee, S. G. & Kim, H. O. Distal-less homeobox 5 is a master regulator of the osteogenesis of human mesenchymal stem cells. Int. J. Mol. Med. 40 , 1486–1494. 10.3892/ijmm.2017.3142 (2017). Gregory, M. D. et al. Sequence variation associated with gene expression is linked to brain structure and function in healthy adults. Cereb. Cortex . 29 , 4654–4661. 10.1093/cercor/bhy344 (2019). Yamazaki, J. et al. Obese status is associated with accelerated DNA methylation change in peripheral blood of senior dogs. Res. Vet. Sci. 139 , 193–199. 10.1016/j.rvsc.2021.07.024 (2021). Main, A. M. et al. DNA methylation and gene expression of HIF3A: cross-tissue validation and associations with BMI and insulin resistance. Clin. Epigenetics . 8 , 89. 10.1186/s13148-016-0258-6 (2016). Benschop, C. C. G. et al. Low template STR typing: Effect of replicate number and consensus method on genotyping reliability and DNA database search results. Forensic Sci. Int-Gen . 5 , 316–328. 10.1016/j.fsigen.2010.06.006 (2011). Shaw, R. J., Akufo-Tetteh, E. K., Risk, J. M., Field, J. K. & Liloglou, T. Methylation enrichment pyrosequencing: combining the specificity of MSP with validation by pyrosequencing. Nucleic Acids Res. 34 , e78. 10.1093/nar/gkl424 (2006). Hayward, K. M. et al. A real-time PCR assay to accurately quantify polar bear DNA in fecal extracts. Peerj 8 , e8884. 10.7717/peerj.8884 (2020). Rallapalli, P. M., Orengo, C. A., Studer, R. A. & Perkins, S. J. Positive selection during the evolution of the blood coagulation factors in the context of their disease-causing mutations. Mol. Biol. Evol. 31 , 3040–3056. 10.1093/molbev/msu248 (2014). Floren, C., Wiedemann, I., Brenig, B., Schütz, E. & Beck, J. Species identification and quantification in meat and meat products using droplet digital PCR (ddPCR). Food Chem. 173 , 1054–1058. 10.1016/j.foodchem.2014.10.138 (2015). Naue, J., Hoefsloot, H. C. J., Kloosterman, A. D. & Verschure, P. J. Forensic DNA methylation profiling from minimal traces: How low can we go? Forensic Sci. Int-Gen . 33 , 17–23. 10.1016/j.fsigen.2017.11.004 (2018). Fan, H. L. et al. Chronological age prediction: developmental evaluation of DNA methylation-based machine learning models. Front. Bioeng. Biotech. 9 , 819991. 10.3389/fbioe.2021.819991 (2022). Xu, C. et al. A novel strategy for forensic age prediction by DNA methylation and support vector regression model. Sci. Rep. 5 , 17788. 10.1038/srep17788 (2015). Monteiro, L. et al. Complex polysaccharides as PCR inhibitors in feces: Helicobacter pylori model. J. Clin. Microbiol. 35 , 995–998. 10.1128/Jcm.35.4.995-998.1997 (1997). Wilson, I. G. Inhibition and facilitation of nucleic acid amplification. Appl. Environ. Microb. 63 , 3741–3751. 10.1128/Aem.63.10.3741-3751.1997 (1997). Ito, T., Katsushima, H., Tomita, K. M. & Matsumoto, T. Infanticide or predation? Cannibalism by a brown bear in Hokkaido. Japan Ursus . 33 , e13. 10.2192/URSUS-D-22-00006.1 (2022). Bellemain, E., Swenson, J. E., Tallmon, O., Brunberg, S. & Taberlet, P. Estimating population size of elusive animals with DNA from hunter-collected feces: Four methods for brown bears. Conserv. Biol. 19 , 150–161. 10.1111/j.1523-1739.2005.00549.x (2005). Robinson, S. J., Waits, L. P. & Martin, I. D. Estimating abundance of American black bears using DNA-based capture-mark-recapture models. Ursus 20 , 1–11. 10.2192/08gr022r.1 (2009). Roda, F. et al. Wolf scat detection dog improves wolf genetic monitoring in new French colonized areas. J. Vertebr Biol. 69 , 20102. 10.25225/jvb.20102 (2021). de Flamingh, A. et al. Combining methods for non-invasive fecal DNA enables whole genome and metagenomic analyses in wildlife biology. Front. Genet. 13 , 1021004. 10.3389/fgene.2022.1021004 (2023). Karmacharya, D. B., Thapa, K., Shrestha, R., Dhakal, M. & Janecka, J. E. Noninvasive genetic population survey of snow leopards ( Panthera uncia ) in Kangchenjunga conservation area, Shey Phoksundo National Park and surrounding buffer zones of Nepal. BMC Res. Notes . 4 , 516. 10.1186/1756-0500-4-516 (2011). Goode, M. J. et al. Capture-recapture of white-tailed deer using DNA from fecal pellet groups. Wildl. Biol. 20 , 270–278. 10.2981/wlb.00050 (2014). Rahman, D. A. et al. The first use of a photogrammetry drone to estimate population abundance and predict age structure of threatened Sumatran elephants. Sci. Rep. 13 , 21311. 10.1038/s41598-023-48635-y (2023). Siren, A. P. K. et al. Monitoring animal populations with cameras using open, multistate, N-mixture models. Ecol. Evol. 14 , e70583. 10.1002/ece3.70583 (2024). Gaillard, J. M., Festa-Bianchet, M. & Yoccoz, N. G. Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends Ecol. Evol. 13 , 58–63. 10.1016/S0169-5347(97)01237-8 (1998). Woodman, J. P. et al. The ecology of ageing in wild societies: linking age structure and social behaviour. Philos. T R Soc. B . 379 , 20220464. 10.1098/rstb.2022.0464 (2024). De Barba, M. et al. DNA metabarcoding multiplexing and validation of data accuracy for diet assessment: application to omnivorous diet. Mol. Ecol. Resour. 14 , 306–323. 10.1111/1755-0998.12188 (2014). Jimbo, M. et al. Diet selection and asocial learning: Natal habitat influence on lifelong foraging strategies in solitary large mammals. Ecosphere 13 , e4105. 10.1002/ecs2.4105 (2022). Riofrío-Lazo, M., Aurioles-Gamboa, D. & Le Boeuf, B. J. Ontogenetic changes in feeding habits of northern elephant seals revealed by δN and δC analysis of growth layers in teeth. Mar. Ecol. Prog Ser. 450 , 229–241. 10.3354/meps09562 (2012). Polischuk, S. C., Hobson, K. A. & Ramsay, M. A. Use of stable-carbon and -nitrogen isotopes to assess weaning and fasting in female polar bears and their cubs. Can. J. Zool. 79 , 499–511. 10.1139/cjz-79-3-499 (2001). Hilderbrand, G. V. et al. The importance of meat, particularly salmon, to body size, population productivity, and conservation of North American brown bears. Can. J. Zool. 77 , 132–138. 10.1139/cjz-77-1-132 (1999). Sommer, F. et al. The gut microbiota modulates energy metabolism in the hibernating brown bear. Cell. Rep. 14 , 1655–1661. 10.1016/j.celrep.2016.01.026 (2016). Ishikawa, A., Sakamoto, H., Katagiri, S. & Takahashi, Y. Changes in sexual behavior and fecal steroid hormone concentrations during the breeding season in female Hokkaido brown bears ( Ursus arctos yesoensis ) under captive condition. J. Vet. Med. Sci. 65 , 99–102. 10.1292/jvms.65.99 (2003). Dalerum, F. et al. Methodological considerations for using fecal glucocorticoid metabolite concentrations as an Indicator of physiological stress in the brown bear ( Ursus arctos ). Physiol. Biochem. Zool. 93 , 227–234. 10.1086/708630 (2020). Dori, P. et al. Hibernating or not hibernating? Brown bears' response to a mismatch between environmental natural cues and captive management, and its welfare implications. Plos One . 19 , e0306537. 10.1371/journal.pone.0306537 (2024). Moriyoshi, M. et al. Patterns of intestinal parasite prevalence in brown bears ( Ursus arctos ) revealed by a 3-year survey on the Shiretoko peninsula, Hokkaido, Japan. Int. J. Parasitol-Par . 26 , 101048. 10.1016/j.ijppaw.2025.101048 (2025). Herrero-García, G. et al. Non-invasive surveillance of shared pathogens in the Eurasian brown bear ( Ursus arctos ) human interface. One Health-Amsterdam . 18 , 100746. 10.1016/j.onehlt.2024.100746 (2024). Shirane, Y. et al. Sex-biased dispersal and inbreeding avoidance in Hokkaido brown bears. J. Mammal . 100 , 1317–1326. 10.1093/jmammal/gyz097 (2019). Zedrosser, A., Stoen, O. G., Sæbo, S. & Swenson, J. E. Should I stay or should I go? Natal dispersal in the brown bear. Anim. Behav. 74 , 369–376. 10.1016/j.anbehav.2006.09.015 (2007). Bojarska, K. & Selva, N. Spatial patterns in brown bear Ursus arctos diet: the role of geographical and environmental factors. Mammal Rev. 42 , 120–143. 10.1111/j.1365-2907.2011.00192.x (2012). Hilderbrand, G. V., Joly, K., Sorum, M. S., Cameron, M. D. & Gustine, D. D. Brown bear ( Ursus arctos ) body size, condition, and productivity in the Arctic, 1977–2016. Polar Biol. 42 , 1125–1130. 10.1007/s00300-019-02501-8 (2019). López-Alfaro, C., Coogan, S. C. P., Robbins, C. T., Fortin, J. K. & Nielsen, S. E. Assessing nutritional parameters of brown bear diets among ecosystems gives insight into differences among populations. Plos One . 10 10.1371/journal.pone.0128088 (2015). Tables Table 1. The genomic location of the targets for methylation level analysis and qPCR, and the genes adjacent to them. Adjacent gene Full name of the gene Number of targets CpG cites DNA positions of target CpG cites in bears (NCBI sequence ID, position) Reference DLX5 Distal-Less Homeobox 5 3 NW_026622985.1, 37441371, 37441374, and 37441376 Tsushima leopard cat 30 SLC12A5 Solute Carrier Family 12 Member 5 4 NW_020656136.1, 29314689, 29314692, 29314694, and 29314698 Brown bear 43 Adjacent gene Full name of the gene qPCR target region Reference F2 Intron12 Coagulation Factor II (F2) NW_020656153.1, 1,273,581–1,273,681 Polar bear 55 Table 2. Values of mean absolute error (MAE), median absolute error (MedAE), and root mean square error (RMSE) after LOOCV for each model. Model Used CpGs* LOOCV model MAE MedAE RMSE Single regression D-1 2.646 1.694 3.843 D-2 2.522 1.767 3.575 D-3 3.100 2.209 4.499 S-1 2.645 2.346 3.794 S-2 2.368 1.477 3.556 S-3 2.330 1.632 3.414 S-4 2.384 1.663 3.358 Elastic net regression D-1, -2 S-1, -2, -3, -4 2.348 1.519 3.465 Support vector regression D-1, -2, -3, S-1, -2, -3, -4 2.163 1.050 3.385 D-1, -2, S-1, -2, -3, -4 2.123 0.984 3.278 D-1, -2, S-2, -3, -4 2.102 0.957 3.267 D-2, S-2, -3, -4 2.085 0.988 3.206 D-2, S-3, -4 2.171 1.136 3.220 D-1, -2, -3 2.493 1.135 3.720 D-1, -2 2.372 1.292 3.550 S-1, -2, -3, -4 2.186 1.763 3.020 S-2, -3, -4 2.109 0.921 3.216 S-3, -4 2.178 1.169 3.199 Note: All values are rounded to the fourth decimal place. *D and S represent DLX5 and SLC12A5, respectively. Each bold value denotes the minimum value for each model. Table 3. Coefficients and p-values for GLMM of Δage in the best model (i.e., SVR model, DLX5-2, SLC12A5-2, -3, and -4). The number of host DNA copies used for analysis means host DNA copy number added to the first PCR. SVR model (DLX5-2, SLC12A5-2, -3, -4) Estimate p-value ΔAge (Intercept) 1.5067 0.1690 Age −0.7271 0.3120 The number of host DNA copies used for analysis −0.3524 0.5540 Environment Wild −1.7127 0.2310 Table 4. Coefficients and p-values for GLMM of |Δage| in the best model (i.e., SVR model, DLX5-2, SLC12A5-2, -3, and -4). SVR model (DLX5-2, SLC12A5-2, -3, -4) Estimate p-value |ΔAge| (Intercept) 2.3049 0.0134 Age 0.9590 0.1050 The number of host DNA copies used for analysis −0.4609 0.0993 Environment Wild −0.1983 0.8592 Additional Declarations No competing interests reported. Supplementary Files SupplementalInformationSciRep.pdf Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 19 May, 2026 Reviewers agreed at journal 27 Apr, 2026 Reviewers agreed at journal 20 Apr, 2026 Reviewers agreed at journal 20 Apr, 2026 Reviews received at journal 12 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers invited by journal 30 Mar, 2026 Editor assigned by journal 27 Mar, 2026 Editor invited by journal 27 Mar, 2026 Submission checks completed at journal 25 Mar, 2026 First submitted to journal 25 Mar, 2026 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-9157516","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":615507923,"identity":"f0b71318-f88a-45c0-9ded-89a074ff76ce","order_by":0,"name":"Satoshi Ohara","email":"","orcid":"","institution":"Hokkaido University","correspondingAuthor":false,"prefix":"","firstName":"Satoshi","middleName":"","lastName":"Ohara","suffix":""},{"id":615507925,"identity":"14f9ac51-05d7-49bc-abe9-6b21848ebfc6","order_by":1,"name":"Shiori Nakamura","email":"","orcid":"","institution":"Hokkaido University","correspondingAuthor":false,"prefix":"","firstName":"Shiori","middleName":"","lastName":"Nakamura","suffix":""},{"id":615507926,"identity":"370a6555-7244-4b9b-ae1c-b231d5b6ec6c","order_by":2,"name":"Kyogo Hagino","email":"","orcid":"","institution":"Noboribetsu Bear Park","correspondingAuthor":false,"prefix":"","firstName":"Kyogo","middleName":"","lastName":"Hagino","suffix":""},{"id":615507927,"identity":"77c74555-4988-4158-ab3b-53a6cedbbfc0","order_by":3,"name":"Yuu Yoshimi","email":"","orcid":"","institution":"Noboribetsu Bear Park","correspondingAuthor":false,"prefix":"","firstName":"Yuu","middleName":"","lastName":"Yoshimi","suffix":""},{"id":615507928,"identity":"3a99a551-2792-4abf-9cd7-ad7433eac08d","order_by":4,"name":"Naoya Matsumoto","email":"","orcid":"","institution":"Azabu University","correspondingAuthor":false,"prefix":"","firstName":"Naoya","middleName":"","lastName":"Matsumoto","suffix":""},{"id":615507930,"identity":"9c38bba7-6681-403e-855f-55e2e817b9a1","order_by":5,"name":"Masami Yamanaka","email":"","orcid":"","institution":"Shiretoko Nature Foundation","correspondingAuthor":false,"prefix":"","firstName":"Masami","middleName":"","lastName":"Yamanaka","suffix":""},{"id":615507933,"identity":"1be26c02-c245-409f-9dac-f223473b737d","order_by":6,"name":"Masanao Nakanishi","email":"","orcid":"","institution":"Shiretoko Nature Foundation","correspondingAuthor":false,"prefix":"","firstName":"Masanao","middleName":"","lastName":"Nakanishi","suffix":""},{"id":615507934,"identity":"04c1a5a0-741b-4397-b66f-c705a0a73506","order_by":7,"name":"Sei Watanabe","email":"","orcid":"","institution":"Shiretoko Nature Foundation","correspondingAuthor":false,"prefix":"","firstName":"Sei","middleName":"","lastName":"Watanabe","suffix":""},{"id":615507935,"identity":"6403a71f-c217-4990-8fd6-ad6ac0e2a4cd","order_by":8,"name":"Hideyuki Ito","email":"","orcid":"","institution":"Kyoto City Zoo","correspondingAuthor":false,"prefix":"","firstName":"Hideyuki","middleName":"","lastName":"Ito","suffix":""},{"id":615507937,"identity":"da0cc1d4-ceb7-441f-90e6-e42a43cae567","order_by":9,"name":"Miho Inoue-Murayama","email":"","orcid":"","institution":"Kyoto University","correspondingAuthor":false,"prefix":"","firstName":"Miho","middleName":"","lastName":"Inoue-Murayama","suffix":""},{"id":615507938,"identity":"95466b63-cdf8-4382-8a64-27730d465466","order_by":10,"name":"Toshio Tsubota","email":"","orcid":"","institution":"Hokkaido University","correspondingAuthor":false,"prefix":"","firstName":"Toshio","middleName":"","lastName":"Tsubota","suffix":""},{"id":615507941,"identity":"dac0dc1e-2672-4ccc-8a57-08578ff64812","order_by":11,"name":"Jumpei Yamazaki","email":"","orcid":"","institution":"Hokkaido University","correspondingAuthor":false,"prefix":"","firstName":"Jumpei","middleName":"","lastName":"Yamazaki","suffix":""},{"id":615507942,"identity":"8fd8d5fd-5c1a-4e50-83e0-a9d4091f348f","order_by":12,"name":"Michito Shimozuru","email":"data:image/png;base64,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","orcid":"","institution":"Hokkaido University","correspondingAuthor":true,"prefix":"","firstName":"Michito","middleName":"","lastName":"Shimozuru","suffix":""}],"badges":[],"createdAt":"2026-03-18 09:38:55","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9157516/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9157516/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105928442,"identity":"2b5f1a0e-53f3-4b22-8910-dc745a5d7a88","added_by":"auto","created_at":"2026-04-01 13:49:50","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":158460,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart illustrates the sample selection process. Numbers represent the count samples at each step, and those in parentheses represent the count of unique individuals. Methylation analysis was performed on 95 samples, and 71 of them were analyzed successfully. Among these, models were constructed using 43 samples and evaluated using 28 samples.\u003c/p\u003e\n\u003cp\u003e*Number of wild individuals after microsatellite analysis included individuals of unknown age.Among these, a portion of the samples from individuals with known actual ages were used.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/9bac3f681aaa81903a2f0059.jpg"},{"id":106959556,"identity":"e691ab85-3e37-4716-bdbc-cfd0673e9ecd","added_by":"auto","created_at":"2026-04-15 09:11:18","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":55174,"visible":true,"origin":"","legend":"\u003cp\u003eScatter plots of age (years) versus DNA methylation level (%) at DLX5-1 (a), -2 (b), -3 (c) SLC12A5-1 (d), -2 (e), -3 (f), and -4 (g) in captive and wild brown bears.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/c37e56265a3c92107b171ed2.jpg"},{"id":105928444,"identity":"2a45b54d-0166-4831-8ef0-5948421655fb","added_by":"auto","created_at":"2026-04-01 13:49:50","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":29669,"visible":true,"origin":"","legend":"\u003cp\u003eScatter plots of predicted age (years) and chronological age (years) for the model with the lowest MAE after LOOCV, for each model type in captive and wild brown bears. The solid line represents predicted age = chronological age. The distance between the dotted line and the solid line represents the MAE of the model after LOOCV:single regression model (a), the best elastic net regression model (b), and the best SVR model (c).\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/94a61366d38aa169d556afd2.jpg"},{"id":105928445,"identity":"2f2fdf8e-7527-49b2-902c-611d3ae7010b","added_by":"auto","created_at":"2026-04-01 13:49:50","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":86698,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of the number of host DNA copies added to the first PCR of methylation analysis between successful and failed samples. Blue: Analysis successful for both \u003cem\u003eDLX5\u003c/em\u003e and \u003cem\u003eSLC12A5\u003c/em\u003e; red: Successful only for \u003cem\u003eDLX5\u003c/em\u003e; green: Successful only for \u003cem\u003eSLC12A5\u003c/em\u003e; purple: Analysis failed for both. In the Success group, the copy number was significantly higher than that in the Failure group (\u003cem\u003eWilcoxon rank-sum test\u003c/em\u003e, p \u0026lt; 0.0001). The dashed line indicates the minimum DNA copy number included in the model construction (i.e., 1,392).\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/684d0a71fee062222aebb94d.jpg"},{"id":106994536,"identity":"0b93c410-ea83-4b74-8c87-3252c6f5e727","added_by":"auto","created_at":"2026-04-15 15:13:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1620130,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/f131ff8b-cf50-4bb9-b077-2f48b821c5d8.pdf"},{"id":105928447,"identity":"0b1246c7-8922-47ba-9999-7e0c0061feeb","added_by":"auto","created_at":"2026-04-01 13:49:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1000206,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalInformationSciRep.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9157516/v1/524dd28bdf633c4979301175.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Establishment of a non-invasive age estimation method based on fecal DNA methylation levels in brown bears Running title: Fecal DNA-based Epigenetic Clock in Brown Bears","fulltext":[{"header":"Introduction","content":"\u003cp\u003eInformation regarding the age of wild animals is crucial for conducting ecological research and developing appropriate conservation and management strategies\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Traditionally, the age of mammals has been estimated by counting the annual growth layers in the cementum of the teeth\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. These layers accumulate with age, and the number of layers is used to estimate chronological age. However, this method has several limitations. First, tooth extraction is required, which is highly invasive in living animals. Furthermore, collecting samples from dead individuals requires specialized skills and considerable effort. Second, the accuracy decreases in older individuals because the cementum layers become increasingly indistinct and difficult to count\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Third, this method requires advanced technical skills and the results may vary depending on the observer\u0026rsquo;s subjective judgment\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Fourth, the formation rate of cementum layers can vary depending on climatic and nutritional conditions, necessitating species- or region-specific calibration\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eOver the last two decades, DNA methylation levels have been used as indicators of age\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. DNA methylation is an epigenetic mechanism that transfers a methyl group to the C5 position of cytosine to form 5-methylcytosine, which primarily occurs on cytosines located within cytosine\u0026ndash;guanine dinucleotide (CpG) sites\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This is a major epigenetic mechanism involved in gene regulation and cellular differentiation. DNA methylation regulates gene expression by preventing transcription factors from binding to DNA or by altering chromatin states to inhibit transcription factor binding\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Previous studies have revealed that the methylation levels at specific CpG sites increase or decrease directionally with age\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, and these age-associated changes have been consistently observed in different tissues and mammalian species\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Statistical models for age prediction (i.e., epigenetic clocks) have been developed based on age-dependent changes in DNA methylation levels\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEpigenetic clock research was initially developed in forensic studies on humans using various biological samples such as blood\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, muscle\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, and saliva\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. In recent years, such studies have been extended to non-human animals. Age estimation models have been established using laboratory animals such as mice (\u003cem\u003eMus musculus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e and naked mole-rats (\u003cem\u003eHeterocephalus glaber\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Similar research has also been conducted on companion and domestic animals, such as dogs (\u003cem\u003eCanis familiaris\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, cats (\u003cem\u003eFelis catus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, and cattle (\u003cem\u003eBos taurus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Moreover, epigenetic clock studies have been performed on various wild species, including humpback whales (\u003cem\u003eMegaptera novaeangliae\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, bottlenose dolphins (\u003cem\u003eTursiops truncatus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, chimpanzees (\u003cem\u003ePan troglodytes\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, wolves (\u003cem\u003eCanis lupus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, and roe deer (\u003cem\u003eCapreolus capreolus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBecause these methods use biological samples such as blood\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, muscle\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, and skin\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, invasive procedures such as capture, anesthesia, and dart biopsy are required. However, in the field, such procedures impose a physical burden on animals and pose potential risks to researchers. Therefore, there is a growing need to develop less invasive or non-invasive sampling methods that can be performed safely with minimal impact on animals. Representative less-invasive or non-invasive samples include hair, feces, and urine, among which feces are particularly easy to collect from many animal species. Feces contain DNA derived from the intestinal epithelial cells of the host\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e, and recent studies have demonstrated that methylation information can also be obtained from fecal DNA. Age estimation based on fecal DNA has been reported for several species, including Indo-Pacific bottlenose dolphins (\u003cem\u003eTursiops aduncus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, wild mice\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, Tsushima leopard cats (\u003cem\u003ePrionailurus bengalensis euptilurus\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. However, studies using fecal DNA are limited, and the target species remain few. The scarcity of methylation studies using fecal DNA may be attributed to several challenges unique to fecal samples. First, collecting feces from free-ranging wild animals of known age across a wide age range, which is essential for constructing an age estimation model, is extremely difficult. Second, fecal DNA is inferior in both quality and quantity compared to DNA obtained from blood or other tissues, and the proportion of host DNA is even lower because it contains DNA derived from intestinal bacteria and food. Third, fecal DNA is highly affected by freshness and environmental exposure, making it difficult to obtain stable and high-quality DNA\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe purpose of this study was to establish an epigenetic clock in feces collected from brown bears (\u003cem\u003eUrsus arctos\u003c/em\u003e). Brown bears have a lifespan of approximately 20\u0026ndash;30 years\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Offspring become independent of their mothers at 1.5\u0026ndash;2.5 years of age\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, and physical growth is completed at approximately 5 years in females and 8 years in males\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. The minimum age at first reproduction is four years\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Hibernation lasts for three to seven months\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e, and pregnant females give birth during hibernation between late January and early February\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Individual and seasonal variations in body size make it almost impossible to identify the age of bears by appearance, except for cubs-of-the-year\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Therefore, tooth-based methods have traditionally been used to estimate the age of brown bears\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Brown bears have low reproductive rates and are vulnerable to overharvesting; thus, population recovery after a decline requires many years\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Human\u0026ndash;bear conflicts, such as crop depredation, intrusion into residential areas, and attacks on livestock and humans, have become serious global issues\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Therefore, understanding the age structure of brown bear populations is essential for developing appropriate conservation and management strategies.\u003c/p\u003e \u003cp\u003eWe previously established an age estimation method for brown bears\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e and other bear species\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e based on DNA methylation levels in the blood; however, this approach requires capture and anesthesia, which are inherently invasive. To address this limitation, a less-invasive technique using hair DNA was recently developed\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. In this study, we aimed to establish a non-invasive and accurate age estimation method for brown bears based on fecal DNA methylation levels.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003eEthical Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll procedures involved in sample collection from live animals in the Shiretoko National Park (incl. a Special Wildlife Protection Area) were conducted in accordance with the Guidelines for Animal Care and Use, Hokkaido University, and were approved by the Animal Care and Use Committee of the Graduate School of Veterinary Medicine, Hokkaido University (permit nos: 1106, 1151, 1152, 15009, 17005, 18-0083, 19-0021, 20-0146, and 23-0014) and by the Hokkaido Regional Environment Office and Kushiro Nature Conservation Office of the Ministry of the Environment, Japan (Permit nos: 2105071, 2205061, 2305161, 2405291, and 2504221). In addition, all methods were carried out in compliance with ARRIVE guidelines.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy area and sampling\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWild bears\u003c/p\u003e\n\u003cp\u003eThe present study was conducted on both wild and captive brown bears. Wild bears were sampled from the Shiretoko Peninsula in eastern Hokkaido, Japan (Supplementary Figure S_M1;\u0026nbsp;43°50′–44°20′\u0026nbsp;N, 144°45′–145°20′\u0026nbsp;E). An area extending from the middle of the peninsula to its tip, covering 610 km², has been designated as a national park. An area covering 711 km², including the national park and the surrounding terrestrial and marine zones, was designated as a UNESCO World Natural Heritage Site in July 2005. Fecal samples collected throughout the peninsula were roughly divided into two groups according to sampling site.\u003c/p\u003e\n\u003cp\u003eThe first sample was collected from the Rusha area of the national park (Supplementary Figure S_M1;\u0026nbsp;44°11′–44°12′\u0026nbsp;N, 145°10′–145°12′\u0026nbsp;E). This area consists of a narrow estuarine coastline approximately 3 km long, and is designated as a Special Wildlife Protection Area, where public access is prohibited without permission. There were no permanent residents except for one fisherman’s settlement. As fishermen have not excluded bears from this area for several decades, they have become habituated to the presence of humans, allowing direct observations at close range. Since the late 1990s, monitoring surveys of identifiable bears based on their external appearance have been conducted and since 2009, DNA samples have been collected from these individuals\u003csup\u003e34,46\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The second set of fecal samples was collected from other areas across the peninsula, including Shari and Rausu towns (Supplementary Figure S_M1). In these areas, management surveys of brown bears have been conducted in response to human-bear conflicts near residential and tourist zones. For some individuals, the actual age was identified through visual observation and individual identification based on DNA analyses\u003csup\u003e46,47\u003c/sup\u003e. Fecal samples collected from these two areas between 2021 and 2025 were used in this study.\u003c/p\u003e\n\u003cp\u003eCaptive bears\u003c/p\u003e\n\u003cp\u003eFecal samples were collected from captive brown bears in July 2025 at Noboribetsu Bear Park (Noboribetsu, Hokkaido, Japan) and the Sapporo Maruyama Zoo (Sapporo, Hokkaido, Japan) (Supplementary Figure S_M1). At the Noboribetsu Bear Park, fecal samples were collected from 31 bears (13 males and 18 females) ranging in age from 0 to 34 years. Among them, two individuals were rescued from the wild as cubs and have since been kept at the facility, whereas all other individuals were born in captivity. When multiple individuals were kept in the same enclosure, multiple feces found within the enclosure were collected. In addition, to ensure sufficient sample numbers, multiple samples were collected from a single fecal deposit. At the Sapporo Maruyama Zoo, two brown bears (a 14-year-old male and a 17-year-old female) were kept, both born in captivity. In the facility, each bear was housed in a separate enclosure. At both facilities, fecal sampling was conducted during routine daily cleaning, ensuring that all collected feces were less than one day old after defecation.\u003c/p\u003e\n\u003cp\u003eSample collection\u003c/p\u003e\n\u003cp\u003eFecal samples were collected according to previously reported methods\u003csup\u003e47\u003c/sup\u003e. For wild bears, only feces estimated to have been defecated within five days were selected, although accurately determining freshness was almost impossible in most cases. The entire surface of each fecal sample was carefully scrubbed using a disposable flocked swab (Flocked Swab R30, Sugiyama-gen Co., ltd., Tokyo, Japan), which was then immersed in a 2.0 mL tube containing 1.1 mL of InhibitEX Buffer (Qiagen inc., Hilden, Germany). This process was repeated several times for each sample. Samples were stored at −20℃ until DNA extraction.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA extraction and individual identification\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDNA was extracted using the QIAamp DNA Stool Mini Kit (Qiagen inc.) according to the manufacturer’s protocol and eluted in 50 µL of elution buffer. To verify the success of DNA extraction and identify individuals, microsatellite analysis was conducted for all fecal DNA samples (i.e., both wild and captive). Based on our previous studies\u003csup\u003e34,47\u003c/sup\u003e, two primer mixtures were used, each containing three microsatellite loci (Primer mix A: G1A, MU05, and MU51; Primer mix B: MU50, G10B, and MU23) and a sex identification marker (Amelogenin)\u003csup\u003e48\u003c/sup\u003e. Polymerase chain reaction (PCR) amplification was performed using a Multiplex PCR Assay Kit (Takara Bio Inc., Shiga, Japan) in a total reaction volume of 10 µL containing 1 µL of genomic DNA, 0.05 µL of Kit Mix 1, 5 µL of Kit Mix 2, 0.5 µL of primer mix (0.25 µmol/L each), and 3.45 µL of PCR-grade water. The PCR conditions were as follows: 95℃ for 30 sec, followed by 40 cycles of 95℃ for 30 sec, 55℃ for 30 sec, and 72℃ for 30 sec, with a final extension at 72℃ for 5 min. The amplified products were analyzed using a SeqStudio Genetic Analyzer (Thermo Fisher Scientific, Waltham, MA, USA) to determine genotypes. As for samples collected in the wild, each identified genotype was compared with previously genotyped individuals from the Shiretoko population to confirm their identity and chronological age, and samples from a subset of these individuals were used. Individual identification was performed for samples collected from enclosures in which multiple bears were housed in captivity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA purification\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAmong the samples with confirmed chronological ages, those that showed stable PCR amplification in microsatellite analysis were purified using the FastGene Gel/PCR Extraction Kit (Nippon Genetics Co., Ltd., Tokyo, Japan). Even if individual identification was possible, samples showing unstable PCR amplification were excluded from subsequent analyses. The purified DNA was eluted in 30–40 µL of elution buffer. In some cases, two samples derived from the same individual collected on the same day were combined before purification to increase the amount of DNA. Subsequently, DNA concentration was measured using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBisulfite conversion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBisulfite treatment was performed using the EZ DNA Methylation-Gold Kit (Zymo Research, Irvine, CA, USA), according to the manufacturer’s protocol. In most cases, 20 µL of DNA solution was added. For samples in which the DNA amount per 20 µL exceeded the maximum processing amount (2,000 ng), the sample was diluted with molecular-grade water to adjust the added DNA amount to 2,000 ng. The volume of elution buffer was adjusted according to the amount of processed DNA to achieve a final DNA concentration of 10 ng/µL.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSelection of target genomic locations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTwo genomic locations adjacent to \u003cem\u003eDLX5\u0026nbsp;\u003c/em\u003e(Distal-less homeobox 5) and \u003cem\u003eSLC12A5\u003c/em\u003e (Solute carrier family 12 member 5) genes (hereafter referred to as \u003cem\u003eDLX5\u0026nbsp;\u003c/em\u003eand\u003cem\u003e\u0026nbsp;SLC12A5\u003c/em\u003e, respectively) were selected as targets for methylation level analysis. The\u003cem\u003e\u0026nbsp;DLX5\u003c/em\u003e is involved in bone development\u003csup\u003e49\u003c/sup\u003e, and CpG sites adjacent to this gene have been reported to correlate with age in multiple felid species\u003csup\u003e30\u003c/sup\u003e. The \u003cem\u003eSLC12A5\u003c/em\u003e encodes the neuron-specific K⁺/Cl⁻ cotransporter KCC2, which mediates rapid hyperpolarization through GABA signaling\u003csup\u003e50\u003c/sup\u003e. The CpG sites adjacent to \u003cem\u003eSLC12A5\u003c/em\u003e in the blood of multiple bear species have been reported to correlate with age\u003csup\u003e43,44\u003c/sup\u003e. For \u003cem\u003eDLX5\u003c/em\u003e, the homologous sequences containing the target CpG sites in the brown bear genome were identified using the Basic Local Alignment Search Tool (BLAST) provided by the National Center for Biotechnology Information (NCBI). For\u0026nbsp;\u003cem\u003eSLC12A5\u003c/em\u003e, the CpG sites reported by Nakamura et al.\u003csup\u003e43\u003c/sup\u003e were used. Detailed information on each genomic region is presented in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePolymerase chain reaction and pyrosequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePCR and subsequent pyrosequencing were conducted as described by Yamazaki et al.\u003csup\u003e51\u003c/sup\u003e. A sample list for the methylation level analysis is shown in Supplementary Table S_M1. PCR was performed using the TaKaRa EpiTaq HS (for bisulfite-treated DNA; Takara Bio Inc.). PCR and pyrosequencing primers were designed using Methyl Primer Express v1.0 (Thermo Fisher Scientific) and PyroMark Assay Design v2.0.2.5 (Qiagen inc.). Details of each primer are listed in Supplementary Table S_M2. PCR was performed in two steps such that biotin-modified primers could be applied to both target genomic locations\u003csup\u003e51\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The first PCR was performed in a total reaction volume of 15 µL containing 1 µL of genomic DNA (diluted to contain 10 ng DNA), 0.075 µL TaKaRa EpiTaq HS, 1.5 µL 10 × EpiTaq PCR Buffer (Mg²⁺-free), 1.5 µL MgCl₂, 1.8 µL dNTP mixture, 0.3 µL each of the forward and reverse primers (10 µmol/L), and 8.525 µL molecular-grade water. The PCR conditions were 98℃ for 1 min, followed by 40 cycles of 98℃ for 10 sec, the annealing temperature (listed in Supplementary Table S_M2) for 30 sec, and 72℃ for 30 sec. Each sample was analyzed in triplicate.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The second PCR was performed in a total reaction volume of 15.13 µL containing 0.1 µL of the first PCR product, 0.075 µL TaKaRa EpiTaq HS, 1.5 µL 10 × EpiTaq PCR Buffer (Mg²⁺- free), 1.5 µL MgCl₂, 1.8 µL dNTP mixture, 0.3 µL of the forward primer (10 µmol/L), 0.06 µL of the reverse primer (10 µmol/L), 0.27 µL of the biotin-modified primer (10 µmol/L), and 9.525 µL molecular-grade water. The PCR conditions were 98℃ for 1 min, followed by 35 cycles of 98℃ for 10 sec, the annealing temperature (listed in Supplementary Table S_M2) for 30 sec, and 72℃ for 30 sec. The first and second PCR products were electrophoresed on a 2% agarose gel to confirm the successful amplification of the target region.\u003c/p\u003e\n\u003cp\u003eThe methylation levels of the target CpG sites (3 and 4 CpGs for \u003cem\u003eDLX1\u003c/em\u003e and \u003cem\u003eSLC12A5\u003c/em\u003e, respectively; Supplementary Figure S_M2) were determined using the PyroMark Q48 software with PyroMark Q48 Advanced Reagents (Qiagen inc.), according to the manufacturer’s protocol. Samples showing unstable baselines in the pyrogram were recommended for reanalysis because the methylation levels could not be correctly calculated\u003csup\u003e52\u003c/sup\u003e. However, because fecal DNA is often of low quality and quantity and to maintain procedural simplicity, these samples were not re-analyzed and were treated as failed analyses. Finally, samples were considered successfully analyzed when at least two of the three replicate measurements satisfied the following two criteria\u003csup\u003e53,54\u003c/sup\u003e: (1) The amplification products of the first and second PCR were confirmed by electrophoresis. 2) In pyrosequencing, the pyrogram showed a stable baseline. For samples that met these criteria, the average methylation level from two or three successful replicates was used as the final methylation level.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDetermination of host DNA copy number using quantitative PCR\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBecause fecal DNA contains not only host DNA, but also intestinal bacteria and dietary DNA, the measured DNA concentration does not necessarily represent the host DNA concentration. Therefore, to quantify the number of host DNA copies used for methylation level analysis, quantitative PCR (qPCR) was performed. The amplification target, the intron 12 region of the coagulation factor II (\u003cem\u003eF2\u003c/em\u003e) gene was selected according to a previous study in which host DNA was quantified using qPCR was developed in polar bears (\u003cem\u003eUrsus maritimus\u003c/em\u003e)\u003csup\u003e55\u003c/sup\u003e. The \u003cem\u003eF2\u003c/em\u003e gene encodes a protein involved in fibrin clot formation\u003csup\u003e56\u003c/sup\u003e, and owing to sequence variations within the intron region, enables species-specific detection\u003csup\u003e57\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Using BLAST provided by NCBI, we confirmed that the sequence of the target region in polar bears (NW_007907185.1; positions 2,099,674–2,099,774) was identical to the homologous sequence in brown bears (NW_020656153.1; positions 1,273,581–1,273,681). Primers with the same sequences described by Hayward et al.\u003csup\u003e55\u003c/sup\u003e (Supplementary Table S_M2) were used. A plasmid containing a 101 bp sequence including the primer attachment region was synthesized (FASMAC, Kanagawa, Japan), and serially diluted to 10⁷, 10⁶, 10⁵, 10⁴, 10³, 10², and 10¹ copies to construct a standard curve. Details of the synthetic plasmids are listed in Supplementary Table S_M3. Standard curves and DNA copy numbers were analyzed using a StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). qPCR was performed using THUNDERBIRD™ SYBR® qPCR Mix (TOYOBO Co., Ltd., Osaka, Japan). Genomic DNA extracted from the blood of two brown bears was included in all qPCR runs as a positive control, and a negative control was included in each run. The PCR was performed in a total reaction volume of 10 µL containing 1 µL of genomic DNA prior to bisulfite conversion, 5 µL THUNDERBIRD SYBR qPCR Mix, 0.2 µL 50×ROX reference dye, 0.2 µL each of the forward and reverse primers (10 µmol/L), and 3.4 µL molecular-grade water. The PCR conditions were 95℃ for 1 min, followed by 40 cycles of 95℃ for 15 sec, 60℃ for 30 sec, and 72℃ for 30 sec, followed by a melting curve analysis (95℃ for 15 sec, 60℃ for 1 min, and 95℃ for 15 sec). Each sample was analyzed in duplicate. The average values of the duplicates were calculated to determine the number of host DNA copies used for the analysis (i.e., the host DNA copy number added to the first PCR).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAge estimation model and model validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo construct a robust age estimation model, only samples meeting the following conditions were used: 1) methylation levels in both the \u003cem\u003eDLX5\u003c/em\u003e and \u003cem\u003eSLC12A5\u003c/em\u003e regions were successfully analyzed, and 2) the number of host DNA copies used for analysis was at least 1,392 or higher. This threshold was based on a previous study\u003csup\u003e58\u003c/sup\u003e reporting that in human methylation analyses, a sample containing ≥ 5 ng host DNA (approximately 1,392 human DNA molecules) is sufficient to obtain stable methylation measurements. If multiple samples from the same individual met these criteria, the sample with the highest number of host DNA copies used for analysis was selected for model construction. In other words, only one sample per individual was used for model construction and the remaining samples were used for model performance evaluation, as described later. Based on the pyrosequencing results, three types of age estimation models were generated following previous studies\u003csup\u003e43-45\u003c/sup\u003e: a single regression model requiring the methylation level of a single CpG site and two multiple regression models (elastic net regression and support vector regression [SVR]) that utilize multiple CpG methylation levels. Age and DNA methylation levels used for model construction were standardized to a mean of 0 and a standard deviation of 1 before integration into the models.\u003c/p\u003e\n\u003cp\u003eSingle regression models were generated using the R command “lm.” Elastic net regression models, a type of penalized regression methods frequently used for age estimation in other species\u003csup\u003e12\u003c/sup\u003e, were constructed using the R package “glmnet.” Optimal parameters (alpha and lambda) were determined using “cv.glmnet”. The SVR models, which have been suggested to outperform elastic net regression for age estimation\u003csup\u003e59,60\u003c/sup\u003e, were constructed using the R package “e1071.” The parameters (cost, gamma, and epsilon) were optimized with the “tune” command under the fixed settings “type = eps-regression, kernel = radial”. CpG site selection for the SVR models was based on the absolute values of regression coefficients (β) obtained from the elastic net regression and was performed by sequentially excluding CpG sites with lower contribution. Additionally, for practical applications, the SVR models using a single region (\u003cem\u003eDLX5\u003c/em\u003e or \u003cem\u003eSLC12A5\u003c/em\u003e) were also constructed. All models were validated using leave-one-out cross-validation (LOOCV), a method that extracts one sample as a test set, while using the remaining samples for training, which is repeated as many times as the number of samples. Model accuracy was evaluated by calculating the mean absolute error (MAE), median absolute error (MedAE), and root mean square error (RMSE) based on the LOOCV results, and the three models were compared.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel performance evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the performance of the constructed models, age estimation was performed using samples that were not included in model construction. These samples included those that met the PCR amplification and pyrogram criteria but had fewer than 1,392 host DNA copies used for analysis, as well as samples not used for model construction because of overlap among individuals. The best-performing model was then used to estimate the ages of the samples not included in model construction, and the estimated ages were compared with the chronological ages.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Generalized linear mixed models (GLMMs) were used to assess whether chronological age, number of host DNA copies used for analysis, and growth environment (i.e., captive or wild) affected the estimation error. Some samples that were not included in the model construction were collected from the same individual. Considering the dependence between samples, individual names were used as random effects. Δage (estimated age − chronological age) and |Δage| (absolute difference between estimated and chronological age) were used as dependent variables, and the three factors (chronological age, the number of host DNA copies used for analysis, and growth environment) and their interactions were included as explanatory variables. Sex was not included as an explanatory variable because older individuals used for model construction were biased toward females (i.e., among individuals aged ≥15 years, five were male and ten were female).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; To evaluate whether the host DNA copy number affected the success of methylation level analysis (i.e., fulfilling the aforementioned criteria in PCR and pyrosequencing), the number of host DNA copies used for analysis was statistically compared between successful and failed cases using the Wilcoxon rank-sum test. The significance level was set at 5%. This comparison included all fecal DNA samples used for the methylation analysis, regardless of whether they were incorporated into the model construction.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eSample selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The samples were selected using the process shown in Figure 1. In total, 25 fecal DNA samples from wild bears (aged 0–28 years) and 18 samples from captive bears (aged 0–26 years) were included in model construction. Additionally, 18 samples from wild bears (10 individuals, aged 1–20 years) and 10 samples from captive bears (8 individuals, aged 1–19 years) were successfully analyzed for DNA methylation levels and were included in the model performance evaluation (i.e., not included in model construction), due to multiple samples from the same individual or insufficient host DNA copy numbers (≤1,392). Detailed information regarding each step is provided in Supplementary Table S_R1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrelation between DNA methylation level and chronological age\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe relationships between DNA methylation levels and the chronological age of the 43 samples included in the model construction are shown in Figure 2. Significant positive correlations were observed for all the CpG sites (Supplementary Figure S_M2; DLX5-1: correlation coefficient [cor] = 0.894, p \u0026lt; 0.001; DLX5-2: cor = 0.909, p \u0026lt; 0.001; DLX5-3: cor = 0.850, p \u0026lt; 0.001; SLC12A5-1: cor = 0.895, p \u0026lt; 0.001; SLC12A5-2: cor = 0.911, p \u0026lt; 0.001; SLC12A5-3: cor = 0.917, p \u0026lt; 0.001; SLC12A5-4: cor = 0.922, p \u0026lt; 0.001). Among these, SLC12A5-4 showed the strongest correlation, although all CpG sites exhibited cor ≥ 0.85.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAge estimation model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eComparison of models\u003c/p\u003e\n\u003cp\u003eBased on the methylation levels at the seven CpG sites, three age estimation models were constructed. The single regression model using the methylation level of SLC12A5-3 showed the highest accuracy, with an MAE of 2.33 years after LOOCV. The formula for age estimation is as follows:\u003c/p\u003e\n\u003cp\u003e“Estimated Age” = (−2.133e−11 + 0.9171 × “methylation level of SLC12A5-3”) × 8.274 (standard deviation of training data) + 10.086 (mean of training data).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The elastic net regression model in which DLX5-3 was excluded because its regression coefficient was zero exhibited the best performance when DLX5-1, DLX5-2, SLC12A5-1, SLC12A5-2, SLC12A5-3, and SLC12A5-4 were used. The MAE after LOOCV is 2.35 years. The formula for age estimation is as follows:\u003c/p\u003e\n\u003cp\u003e“Estimated Age” = (2.32e−11 + 0.1033 × “methylation level of DLX5-1” + 0.1577 × “methylation level of DLX5-2” + 0.0990 × “methylation level of SLC12A5-1” + 0.1414 × “methylation level of SLC12A5-2” + 0.1628 × “methylation level of SLC12A5-3” + 0.2348 × “methylation level of SLC12A5-4”) × 8.274 + 10.086.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;The support vector regression (SVR) models using DLX5-2, SLC12A5-2, SLC12A5-3, and SLC12A5-4 exhibited the highest accuracy, with an MAE of 2.08 years after LOOCV. The detailed statistics of each model are presented in Table 2, and the relationships between the chronological and estimated ages after LOOCV are shown in Figure 3. The regression coefficients of the elastic net model and the parameter settings for the elastic net regression and SVR models are summarized in Supplementary Tables S_R2 and S_R3. The R script used for age estimation is provided in the Supplementary Materials.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel performance evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; The SVR model using DLX5-2 and SLC12A5-2, -3, and -4, which exhibited the lowest MAE after LOOCV, was selected as the best model. When applied to samples that were not included in the model construction, the model achieved an MAE of 2.10, a MedAE of 1.35, and an RMSE of 2.93. The relationship between chronological and estimated ages and the results of the generalized linear mixed models (GLMMs) are shown in Supplementary Figure S_R1 and Table 3–4, respectively. Neither chronological age nor growth environment had a significant effect on either Δage (predicted age − chronological age) or |Δage| (absolute difference between predicted and chronological age). Samples with lower host copy numbers used for analysis tended to show larger prediction errors (i.e., |Δage|), although they were not statistically significant (p = 0.09). None of the interactions among the explanatory variables were statistically significant. In addition, samples in which the methylation level analysis failed had significantly lower host DNA copy numbers used for analysis than successful samples (p \u0026lt; 0.001; Figure 4).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis is the first study to establish a non-invasive method for epigenetic age estimation using fecal DNA from bears. Based on the MAE after LOOCV, it was suggested that the best model was the SVR model constructed using four CpG sites (i.e., DLX5-2 and SLC12A5-2, -3, and -4) with an MAE of 2.08 years (6.9\u0026ndash;10.4% of lifespan). This accuracy was slightly inferior to the age estimation models developed using blood DNA of brown bears (MAE: 1.30; 4.3\u0026ndash;6.5% of lifespan\u003csup\u003e43\u003c/sup\u003e), but superior to the hair DNA-based model (MAE: 3.19; 10.6\u0026ndash;16.0% of lifespan\u003csup\u003e45\u003c/sup\u003e). In previous studies using fecal DNA, the reported MAE values were 5.08 years for Indo-Pacific bottlenose dolphins (lifespan: approx. 40\u0026ndash;50 years; about 10.2\u0026ndash;12.7% of lifespan\u003csup\u003e28\u003c/sup\u003e), 2.54 years for Tsushima leopard cats (lifespan: 15\u0026ndash;20 years; 12.7\u0026ndash;16.9% of lifespan\u003csup\u003e30\u003c/sup\u003e), 26 days for mice (lifespan: 30\u0026ndash;32 months; 3% of lifespan\u003csup\u003e29\u003c/sup\u003e). Compared to these species, the current model achieved comparable or higher accuracy for brown bears (lifespan: 20\u0026ndash;30 years\u003csup\u003e33\u003c/sup\u003e). When comparing the SVR models constructed using only\u003cem\u003e\u0026nbsp;SLC12A5\u0026nbsp;\u003c/em\u003e(SLC12A5-2, -3, and -4) to the best model, the difference in MAE was negligible (2.11 vs. 2.08 years) and MedAE values were slightly superior in the former case (0.92 vs. 0.99 years). Therefore, when analyzing many samples, a model constructed using three CpG sites in \u003cem\u003eSLC12A5\u003c/em\u003e, which requires the amplification of only one genomic region, is more practical and cost-effective. The CpG sites adjacent to\u0026nbsp;\u003cem\u003eSLC12A5\u003c/em\u003e have also been used as a single-target locus in an age estimation model for blood DNA\u003csup\u003e43,44\u003c/sup\u003e and as the best-performing target in a fecal DNA-based model for Tsushima leopard cats\u003csup\u003e30\u003c/sup\u003e. These results suggest\u0026nbsp;that this genomic location may serve as a valuable target for developing DNA methylation-based age estimation models across different sample types and mammalian species.\u003c/p\u003e\n\u003cp\u003eAge estimation using fecal DNA has several challenges, including the presence of PCR inhibitors in feces that can hinder DNA amplification\u003csup\u003e61,62\u003c/sup\u003e, unavoidable contamination by DNA derived from diet and intestinal bacteria, and degradation of host DNA quantity and quality due to environmental exposure\u003csup\u003e31,32\u003c/sup\u003e. In this study, we purified all samples to minimize the inclusion of inhibitory components and remove short, fragmented DNA. Furthermore, quantification of the host DNA copy numbers used for the analysis clarified their impact on the success or failure of methylation-level measurements. These newly added steps were not present in previous fecal DNA-based studies\u003csup\u003e28-30\u003c/sup\u003e and significantly contributed to the success of the analysis and improvement of its accuracy. The number of host DNA copies used for analysis is a clear factor that determines the success or failure of the methylation-level analysis. In addition, there was a tendency that the estimation error (i.e., |\u0026Delta;age|) to increase as the host DNA copy number decreased. This suggests that for biological samples, such as fecal DNA, which often present challenges in DNA quantity and quality, pre-quantification of host DNA copy number may enable prior decisions on sample suitability for analysis, thereby enhancing cost-effectiveness in practical applications. In the current study\u0026rsquo;s age estimation model construction, a threshold of 1,392 host DNA copies used for analysis was set to ensure model robustness\u003csup\u003e58\u003c/sup\u003e. Nevertheless, among the 30 samples that did not meet this criterion, methylation level analysis was successful in 10 samples (the lowest host DNA copy number used for analysis among the successful samples was 172). This indicates that even samples below the threshold can be used for methylation level estimation if it is possible to include approximately 150 host DNA copies or more in the first PCR and successful amplification is confirmed by electrophoresis. However, it should be kept in mind that there is an increased risk of estimation error with a lower host DNA copy number. Intron 12 of the\u0026nbsp;\u003cem\u003eF2\u003c/em\u003e gene, targeted for qPCR quantification of host DNA, has been shown to be unaffected by DNA derived from prey species of polar bears, including the American moose (\u003cem\u003eAlces alces\u003c/em\u003e), Arctic hares (\u003cem\u003eLepus arcticus\u003c/em\u003e), spotted seals (\u003cem\u003ePhoca largha\u003c/em\u003e), ringed seals (\u003cem\u003ePusa hispida\u003c/em\u003e), and Arctic foxes (\u003cem\u003eVulpes lagopus\u003c/em\u003e), and\u0026nbsp;is not affected by\u0026nbsp;contamination with human DNA\u003csup\u003e55\u003c/sup\u003e. Nonetheless, because cannibalism, including infanticide, has been documented in brown bears\u003csup\u003e63\u003c/sup\u003e, the present method may not be applicable to feces containing bear carcasses.\u003c/p\u003e\n\u003cp\u003eThe use of fecal DNA for age estimation is highly applicable. One major advantage is the potential to improve the accuracy of population age structure and population dynamic assessments. Population monitoring of bears using fecal samples has been conducted worldwide\u003csup\u003e64,65\u003c/sup\u003e. Similar studies have been conducted on other wildlife species, including wolves\u003csup\u003e66\u003c/sup\u003e, elephants (\u003cem\u003eLoxodonta Africana\u003c/em\u003e)\u003csup\u003e67\u003c/sup\u003e, snow leopards (\u003cem\u003ePanthera uncia\u003c/em\u003e)\u003csup\u003e68\u003c/sup\u003e, and white-tailed deer (\u003cem\u003eOdocoileus virginianus\u003c/em\u003e)\u003csup\u003e69\u003c/sup\u003e. These studies used DNA information obtained from samples to infer population dynamics using capture-mark-recapture methods or spatial capture-recapture approaches. However, DNA information alone does not provide the age structure of a population, requiring additional methods, such as camera trapping or direct visual surveys\u003csup\u003e70,71\u003c/sup\u003e. By applying the method established in this study, the age structure of a population can be inferred from fecal samples. It is essential to identify age classes such as juvenile and reproductively mature individuals to make more accurate predictions of future population trends\u003csup\u003e72,73\u003c/sup\u003e. This study provides valuable information for population management and conservation.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Feces also serve as biological samples that provide a wide range of valuable information for understanding the life histories of animals. A representative example of brown bears is dietary information\u003csup\u003e39,74\u003c/sup\u003e. These studies analyzed diet using DNA barcoding and identified food remains in fecal samples. Although such approaches can reveal regional and seasonal differences in diet, age-related dietary changes have relied on indirect methods such as stable isotope analyses using biological samples (e.g., hair\u003csup\u003e75\u003c/sup\u003e, teeth\u003csup\u003e76\u003c/sup\u003e, and blood\u003csup\u003e77\u003c/sup\u003e). By combining fecal content analysis with age estimation, it is possible to clarify age-related shifts in resource use and foraging strategies, allowing for a deeper understanding of the life history of brown bears. In addition, determining age-specific dependence on resources can help identify key environments that must be conserved to maintain population size, thereby making a substantial contribution to wildlife conservation\u003csup\u003e78\u003c/sup\u003e. Similarly, other applications are expected, such as studies on the gut microbiome\u003csup\u003e79\u003c/sup\u003e, analyses of fecal hormones (e.g., fecal estradiol and progesterone\u003csup\u003e80\u003c/sup\u003e and fecal glucocorticoids\u003csup\u003e81,82\u003c/sup\u003e), research on gastrointestinal parasites\u003csup\u003e83\u003c/sup\u003e, and infectious diseases\u003csup\u003e84\u003c/sup\u003e. The methodology developed in this study elucidates age-dependent biological changes, including nutrition, reproduction, stress, disease infection, and immunity. This study provides valuable insights into the adaptive strategies and physiological changes in wild populations.\u003c/p\u003e\n\u003cp\u003eOne of the limitations of this study was that sex differences could not be fully evaluated because of sampling bias. In the samples used for model construction, older individuals were biased toward females (two-thirds of the individuals aged \u0026ge;15 years were females). The shortage of samples from adult males, particularly wild males, was notable; only five males were included in the model construction, and all were 4 years old or younger. Wild male bears do not remain in their natal areas and typically disperse at approximately three years of age, with dispersal distances longer than those of females\u003csup\u003e85,86\u003c/sup\u003e. Therefore, it was not possible to collect fecal samples from adult males born within the study area whose ages were known.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Another challenge that needs to be addressed is determining whether the present model can be applied to other brown bear populations. The samples used for the model construction were limited to wild individuals from a restricted area in Hokkaido and two captive facilities, all of which were one subspecies of the brown bear. Brown bears are among the most widely distributed terrestrial mammalian species, occupying broad regions across North America, Europe, and Asia. They inhabit diverse environments, including forests, coastal zones, savannas, desert areas, and alpine zones, and exhibit substantial regional variations in diet and body size\u003csup\u003e87,88\u003c/sup\u003e. For example, a study of three North American populations showed that, even within North America, coastal and inland groups differ in the principal food items that influence body size\u003csup\u003e89\u003c/sup\u003e. Although this study found no effect of environment (i.e., captive or wild) on the estimation error, it remains necessary to verify whether the current age estimation model can be applied to other brown bear populations without calibration.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;In conclusion, this study established the first epigenetic clock for brown bears using fecal DNA samples based on the methylation levels of CpG sites adjacent to the\u003cem\u003e\u0026nbsp;DLX5\u003c/em\u003e and \u003cem\u003eSLC12A5\u003c/em\u003e genes. The best model achieved an MAE of 2.08 years, demonstrating high accuracy. Fecal samples can be collected non-invasively, which is a clear distinction from existing techniques for estimating the age of brown bears and other species based on blood, hair, and other sample types.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eFurthermore, a new process was developed that enables the preliminary selection and precise evaluation of samples by quantifying the number of host DNA copies present in fecal DNA. This approach addresses issues specific to fecal DNA, such as contamination by nonhost DNA and degradation of DNA quality due to environmental exposure, which are essential considerations for practical applications in field research. This study holds substantial value for ecological research and conservation of brown bears and is expected to contribute to the development of non-invasive approaches for other species.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our profound gratitude and appreciation to all the people who cooperated with the sampling, especially to all the staff of the Shiretoko Nature Foundation, Noboribetsu bear park, and Sapporo Maruyama Zoo (especially to Mr. Wataru Goshima and Mr. Michiaki Shimizu) for providing bear fecal samples.\u0026nbsp;We wish to thank Hatsusaburo Ose and all the members of the Shiretoko Fishery Productive Association for their kind support. Finally, we thank Editage (www.editage.jp) for English language editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by funding from the Japan Society for the Promotion of Science (JSPS) (https://www.jsps.go.jp/english/e-grants/index.html) KAKENHI (grant numbers: JP19K06833, JP22K14910, JP23K05312, JP24KJ0304, JP25K22873 and JP25H01002), the Collaborative Research Program of Wildlife Research Center, Kyoto University (grant no. 2021-A-16), and Grant for Basic Science Research Projects from The Sumitomo Foundation (grant no. 200561).\u0026nbsp;This research was performed by the Environment Research and Technology Development Fund (JPMEERF20254002) of the Environmental Restoration and Conservation Agency provided by Ministry of the Environment of Japan.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.O., S.N., J.Y. and M.S. designed the study. S.O., S.N., and M.S. performed laboratory work, and constructed each age estimation model. S.O. S.N., K.H., Y.Y., N.M., M.Y., M.N., S.W., and M.S. were involved in sample collection. J.Y. and H.I., and M. I-M. supported the technical aspects of the experiment. S.O. and M.S. wrote the article with inputs from S.N., J.Y., H.I., and T.T. All authors reviewed the article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets and R-scripts used in this study are available from the corresponding authors on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eColchero, F. et al. The diversity of population responses to environmental change. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 342\u0026ndash;353. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ele.13195\u003c/span\u003e\u003cspan address=\"10.1111/ele.13195\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobert, A., Chantepie, S., Pavard, S., Sarrazin, F. \u0026amp; Teplitsky, C. Actuarial senescence can increase the risk of extinction of mammal populations. \u003cem\u003eEcol. Appl.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 116\u0026ndash;124. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1890/14-0221.1\u003c/span\u003e\u003cspan address=\"10.1890/14-0221.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTidi\u0026egrave;re, M. et al. Comparative analyses of longevity and senescence reveal variable survival benefits of living in zoos across mammals. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 36361. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/srep36361\u003c/span\u003e\u003cspan address=\"10.1038/srep36361\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThomas, D. C. Metachromatic staining of dental cementum for mammalian age-determination. \u003cem\u003eJ. Wildl. Manage.\u003c/em\u003e \u003cb\u003e41\u003c/b\u003e, 207\u0026ndash;210 (1977). doi:Doi 10.2307/3800596.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcLaughlin, C. R. et al. Precision and accuracy of estimating age of maine black bears by cementum annuli. \u003cem\u003eInt. Conf. Bear. Res. Manage.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 415\u0026ndash;419. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2307/3872945\u003c/span\u003e\u003cspan address=\"10.2307/3872945\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1990).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRolandsen, C. M. et al. Accuracy and repeatability of moose (\u003cem\u003eAlces alces\u003c/em\u003e) age as estimated from dental cement layers. \u003cem\u003eEur. J. Wildl. Res.\u003c/em\u003e \u003cb\u003e54\u003c/b\u003e, 6\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10344-007-0100-8\u003c/span\u003e\u003cspan address=\"10.1007/s10344-007-0100-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBocklandt, S. et al. Epigenetic predictor of age. \u003cem\u003ePlos One\u003c/em\u003e. \u003cb\u003e6\u003c/b\u003e, e14821. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0014821\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0014821\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBogdanovic, O. \u0026amp; Veenstra, G. J. C. DNA methylation and methyl-CpG binding proteins: developmental requirements and function. \u003cem\u003eChromosoma\u003c/em\u003e \u003cb\u003e118\u003c/b\u003e, 549\u0026ndash;565. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00412-009-0221-9\u003c/span\u003e\u003cspan address=\"10.1007/s00412-009-0221-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore, L. D., Le, T. \u0026amp; Fan, G. P. DNA methylation and its basic function. \u003cem\u003eNeuropsychopharmacol\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 23\u0026ndash;38. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/npp.2012.112\u003c/span\u003e\u003cspan address=\"10.1038/npp.2012.112\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, M. J., Goodman, S. J. \u0026amp; Kobor, M. S. DNA methylation and healthy human aging. \u003cem\u003eAging Cell.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 924\u0026ndash;932. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/acel.12349\u003c/span\u003e\u003cspan address=\"10.1111/acel.12349\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu, A. T. et al. Universal DNA methylation age across mammalian tissues. \u003cem\u003eNat. Aging\u003c/em\u003e. \u003cb\u003e3\u003c/b\u003e, 1462\u0026ndash;1462. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s43587-023-00499-7\u003c/span\u003e\u003cspan address=\"10.1038/s43587-023-00499-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorvath, S. DNA methylation age of human tissues and cell types. \u003cem\u003eGenome Biol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, R115. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/gb-2013-14-10-r115\u003c/span\u003e\u003cspan address=\"10.1186/gb-2013-14-10-r115\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBekaert, B., Kamalandua, A., Zapico, S. C., Van de Voorde, W. \u0026amp; Decorte, R. Improved age determination of blood and teeth samples using a selected set of DNA methylation markers. \u003cem\u003eEpigenetics\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 922\u0026ndash;930. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/15592294.2015.1080413\u003c/span\u003e\u003cspan address=\"10.1080/15592294.2015.1080413\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVoisin, S. et al. An epigenetic clock for human skeletal muscle. \u003cem\u003eJ. Cachexia Sarcopenia Muscle\u003c/em\u003e. \u003cb\u003e11\u003c/b\u003e, 887\u0026ndash;898. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jcsm.12556\u003c/span\u003e\u003cspan address=\"10.1002/jcsm.12556\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong, S. R. et al. DNA methylation-based age prediction from saliva: High age predictability by combination of 7 CpG markers. \u003cem\u003eForensic Sci. International: Genet.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 118\u0026ndash;125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.fsigen.2017.04.006\u003c/span\u003e\u003cspan address=\"10.1016/j.fsigen.2017.04.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStubbs, T. M. et al. Multi-tissue DNA methylation age predictor in mouse. \u003cem\u003eGenome Biol.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e, 68. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13059-017-1203-5\u003c/span\u003e\u003cspan address=\"10.1186/s13059-017-1203-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLowe, R. et al. DNA methylation clocks as a predictor for ageing and age estimation in naked mole-rats,. \u003cem\u003eAging\u003c/em\u003e 12, 4394\u0026ndash;4406, (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18632/aging.102892\u003c/span\u003e\u003cspan address=\"10.18632/aging.102892\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson, M. J., Vonholdt, B., Horvath, S. \u0026amp; Pellegrini, M. An epigenetic aging clock for dogs and wolves. \u003cem\u003eAging\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 1055\u0026ndash;1068. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18632/aging.101211\u003c/span\u003e\u003cspan address=\"10.18632/aging.101211\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi, H. et al. Age estimation using methylation-sensitive high-resolution melting (MS-HRM) in both healthy felines and those with chronic kidney disease. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 19963. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-021-99424-4\u003c/span\u003e\u003cspan address=\"10.1038/s41598-021-99424-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaj, K. et al. Epigenetic clock and methylation studies in cats. \u003cem\u003eGeroscience\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 2363\u0026ndash;2378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11357-021-00445-8\u003c/span\u003e\u003cspan address=\"10.1007/s11357-021-00445-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaulton, A. et al. Development of epigenetic clocks for key ruminant species. \u003cem\u003eGenes\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/genes13010096\u003c/span\u003e\u003cspan address=\"10.3390/genes13010096\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePolanowski, A. M., Robbins, J., Chandler, D. \u0026amp; Jarman, S. N. Epigenetic estimation of age in humpback whales. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 976\u0026ndash;987. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1755-0998.12247\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.12247\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeal, A. P., Kiszka, J. J., Wells, R. S. \u0026amp; Eirin-Lopez, J. M. The bottlenose dolphin epigenetic aging tool (BEAT): A molecular age estimation tool for small cetaceans. \u003cem\u003eFront. Mar. Sci.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 561. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fmars.2019.00561\u003c/span\u003e\u003cspan address=\"10.3389/fmars.2019.00561\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIto, H., Udono, T., Hirata, S. \u0026amp; Inoue-Murayama, M. Estimation of chimpanzee age based on DNA methylation. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 9998. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-018-28318-9\u003c/span\u003e\u003cspan address=\"10.1038/s41598-018-28318-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLemaitre, J. F. et al. DNA methylation as a tool to explore ageing in wild roe deer populations. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 1002\u0026ndash;1015. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1755-0998.13533\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.13533\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePetkovich, D. A. et al. Using DNA methylation profiling to evaluate biological age and longevity interventions. \u003cem\u003eCell. Metab.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 954\u0026ndash;960e956. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cmet.2017.03.016\u003c/span\u003e\u003cspan address=\"10.1016/j.cmet.2017.03.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarrero, P., Fregel, R., Cabrera, V. M. \u0026amp; Nogales, M. Extraction of high-quality host DNA from feces and regurgitated seeds: a useful tool for vertebrate ecological studies. \u003cem\u003eBiol. Res.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 147\u0026ndash;151 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYagi, G. et al. Non-invasive age estimation based on faecal DNA using methylation-sensitive high-resolution melting for Indo-Pacific bottlenose dolphins. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, e13906. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1755-0998.13906\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.13906\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanski, E. et al. Epigenetic age estimation of wild mice using faecal samples. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, e17330. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/mec.17330\u003c/span\u003e\u003cspan address=\"10.1111/mec.17330\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi, H. \u003cem\u003eDevelopment of epigenetic clocks in multiple felid species \u0026mdash; from small to big, domestic to wild.\u003c/em\u003e Doctral thesis, Kyoto University, (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.14989/doctor.k25151\u003c/span\u003e\u003cspan address=\"10.14989/doctor.k25151\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgetsuma-Yanagihara, Y., Inoue, E. \u0026amp; Agetsuma, N. Effects of time and environmental conditions on the quality of DNA extracted from fecal samples for genotyping of wild deer in a warm temperate broad-leaved forest. \u003cem\u003eMammal Res.\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e, 201\u0026ndash;207. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s13364-016-0305-x\u003c/span\u003e\u003cspan address=\"10.1007/s13364-016-0305-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrinkman, T. J., Schwartz, M. K., Person, D. K., Pilgrim, K. L. \u0026amp; Hundertmark, K. J. Effects of time and rainfall on PCR success using DNA extracted from deer fecal pellets. \u003cem\u003eConserv. Genet.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 1547\u0026ndash;1552. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10592-009-9928-7\u003c/span\u003e\u003cspan address=\"10.1007/s10592-009-9928-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBunnell, F. L. \u0026amp; Trait, D. E. N. in \u003cem\u003eDynamics of Large Mammal Populations\u003c/em\u003e (eds C. W. Fowler \u0026amp; T. D. Smith) Ch. 4, 75\u0026ndash;98 (Wiley-Interscience, (1981).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimozuru, M. et al. Reproductive parameters and cub survival of brown bears in the Rusha area of the Shiretoko Peninsula, Hokkaido, Japan. \u003cem\u003ePlos One\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, e0176251. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0176251\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0176251\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShirane, Y. et al. Development of a noninvasive photograph-based method for the evaluation of body condition in free-ranging brown bears. \u003cem\u003ePeerj\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, e9982. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7717/peerj.9982\u003c/span\u003e\u003cspan address=\"10.7717/peerj.9982\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMano, T. \u0026amp; Tsubota, T. Reproductive characteristics of brown bears on the Oshima Peninsula, Hokkaido, Japan. \u003cem\u003eJ. Mammal\u003c/em\u003e. \u003cb\u003e83\u003c/b\u003e, 1026\u0026ndash;1034. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1644/1545-1542(2002)083%3C1026\u003c/span\u003e\u003cspan address=\"10.1644/1545-1542(2002)083%3C1026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2002). :Rcobbo\u0026gt;2.0.Co;2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Bernardo, E., Russo, L. F., Valderr\u0026aacute;bano, E., Fern\u0026aacute;ndez, A. \u0026amp; Penteriani, V. Denning in brown bears. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 6844\u0026ndash;6862. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ece3.6372\u003c/span\u003e\u003cspan address=\"10.1002/ece3.6372\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriebe, A. et al. Factors affecting date of implantation, parturition, and den entry estimated from activity and body temperature in free-ranging brown bears. \u003cem\u003ePlos One\u003c/em\u003e. \u003cb\u003e9\u003c/b\u003e, e101410. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0101410\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0101410\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShirane, Y. et al. Dining from the coast to the summit: Salmon and pine nuts determine the summer body condition of female brown bears on the Shiretoko Peninsula. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 5204\u0026ndash;5219. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ece3.7410\u003c/span\u003e\u003cspan address=\"10.1002/ece3.7410\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiller, S. D. Population management of bears in North America. \u003cem\u003eBears: Their Biology Manage.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 357\u0026ndash;373. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2307/3872940\u003c/span\u003e\u003cspan address=\"10.2307/3872940\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1990).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBombieri, G. et al. Brown bear attacks on humans: a worldwide perspective. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 8573. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-019-44341-w\u003c/span\u003e\u003cspan address=\"10.1038/s41598-019-44341-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCan, \u0026Ouml;. E., D'Cruze, N., Garshelis, D. L., Beecham, J. \u0026amp; Macdonald, D. W. Resolving human-bear conflict: A global survey of countries, experts, and key factors. \u003cem\u003eConserv. Lett.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, 501\u0026ndash;513. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/conl.12117\u003c/span\u003e\u003cspan address=\"10.1111/conl.12117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakamura, S. et al. Age estimation based on blood DNA methylation levels in brown bears. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 1211\u0026ndash;1225. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1755-0998.13788\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.13788\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimozuru, M. et al. Epigenetic clock in bears: A simple cost-effective blood DNA methylation-based age estimation method applicable to multiple bear species. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, e71424. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ece3.71424\u003c/span\u003e\u003cspan address=\"10.1002/ece3.71424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakamura, S. et al. Less-invasive age estimation using hair based on DNA methylation in brown bears. \u003cem\u003eSci. Rep-Uk\u003c/em\u003e. \u003cb\u003e15\u003c/b\u003e, 45696. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-025-27455-2\u003c/span\u003e\u003cspan address=\"10.1038/s41598-025-27455-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimozuru, M. et al. Incidence of multiple paternity and inbreeding in high-density brown bear populations on the Shiretoko Peninsula, Hokkaido, Japan. \u003cem\u003eJ. Hered\u003c/em\u003e. \u003cb\u003e110\u003c/b\u003e, 321\u0026ndash;331. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/jhered/esz002\u003c/span\u003e\u003cspan address=\"10.1093/jhered/esz002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimozuru, M. et al. Estimation of breeding population size using DNA-based pedigree reconstruction in brown bears. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, e9246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ece3.9246\u003c/span\u003e\u003cspan address=\"10.1002/ece3.9246\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYamamoto, K. et al. Sex identification of Japanese black bear, \u003cem\u003eUrsus thibetanus japonicus\u003c/em\u003e, by PCR based on amelogenin gene. \u003cem\u003eJ. Vet. Med. Sci.\u003c/em\u003e \u003cb\u003e64\u003c/b\u003e, 505\u0026ndash;508. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1292/jvms.64.505\u003c/span\u003e\u003cspan address=\"10.1292/jvms.64.505\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeo, J. S., Lee, S. G. \u0026amp; Kim, H. O. Distal-less homeobox 5 is a master regulator of the osteogenesis of human mesenchymal stem cells. \u003cem\u003eInt. J. Mol. Med.\u003c/em\u003e \u003cb\u003e40\u003c/b\u003e, 1486\u0026ndash;1494. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3892/ijmm.2017.3142\u003c/span\u003e\u003cspan address=\"10.3892/ijmm.2017.3142\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGregory, M. D. et al. Sequence variation associated with gene expression is linked to brain structure and function in healthy adults. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e29\u003c/b\u003e, 4654\u0026ndash;4661. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/cercor/bhy344\u003c/span\u003e\u003cspan address=\"10.1093/cercor/bhy344\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYamazaki, J. et al. Obese status is associated with accelerated DNA methylation change in peripheral blood of senior dogs. \u003cem\u003eRes. Vet. Sci.\u003c/em\u003e \u003cb\u003e139\u003c/b\u003e, 193\u0026ndash;199. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.rvsc.2021.07.024\u003c/span\u003e\u003cspan address=\"10.1016/j.rvsc.2021.07.024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMain, A. M. et al. DNA methylation and gene expression of HIF3A: cross-tissue validation and associations with BMI and insulin resistance. \u003cem\u003eClin. Epigenetics\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, 89. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13148-016-0258-6\u003c/span\u003e\u003cspan address=\"10.1186/s13148-016-0258-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenschop, C. C. G. et al. Low template STR typing: Effect of replicate number and consensus method on genotyping reliability and DNA database search results. \u003cem\u003eForensic Sci. Int-Gen\u003c/em\u003e. \u003cb\u003e5\u003c/b\u003e, 316\u0026ndash;328. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.fsigen.2010.06.006\u003c/span\u003e\u003cspan address=\"10.1016/j.fsigen.2010.06.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShaw, R. J., Akufo-Tetteh, E. K., Risk, J. M., Field, J. K. \u0026amp; Liloglou, T. Methylation enrichment pyrosequencing: combining the specificity of MSP with validation by pyrosequencing. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e, e78. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gkl424\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkl424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHayward, K. M. et al. A real-time PCR assay to accurately quantify polar bear DNA in fecal extracts. \u003cem\u003ePeerj\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, e8884. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7717/peerj.8884\u003c/span\u003e\u003cspan address=\"10.7717/peerj.8884\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRallapalli, P. M., Orengo, C. A., Studer, R. A. \u0026amp; Perkins, S. J. Positive selection during the evolution of the blood coagulation factors in the context of their disease-causing mutations. \u003cem\u003eMol. Biol. Evol.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 3040\u0026ndash;3056. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/molbev/msu248\u003c/span\u003e\u003cspan address=\"10.1093/molbev/msu248\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFloren, C., Wiedemann, I., Brenig, B., Sch\u0026uuml;tz, E. \u0026amp; Beck, J. Species identification and quantification in meat and meat products using droplet digital PCR (ddPCR). \u003cem\u003eFood Chem.\u003c/em\u003e \u003cb\u003e173\u003c/b\u003e, 1054\u0026ndash;1058. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.foodchem.2014.10.138\u003c/span\u003e\u003cspan address=\"10.1016/j.foodchem.2014.10.138\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaue, J., Hoefsloot, H. C. J., Kloosterman, A. D. \u0026amp; Verschure, P. J. Forensic DNA methylation profiling from minimal traces: How low can we go? \u003cem\u003eForensic Sci. Int-Gen\u003c/em\u003e. \u003cb\u003e33\u003c/b\u003e, 17\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.fsigen.2017.11.004\u003c/span\u003e\u003cspan address=\"10.1016/j.fsigen.2017.11.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan, H. L. et al. Chronological age prediction: developmental evaluation of DNA methylation-based machine learning models. \u003cem\u003eFront. Bioeng. Biotech.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 819991. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fbioe.2021.819991\u003c/span\u003e\u003cspan address=\"10.3389/fbioe.2021.819991\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, C. et al. A novel strategy for forensic age prediction by DNA methylation and support vector regression model. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, 17788. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/srep17788\u003c/span\u003e\u003cspan address=\"10.1038/srep17788\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonteiro, L. et al. Complex polysaccharides as PCR inhibitors in feces: Helicobacter pylori model. \u003cem\u003eJ. Clin. Microbiol.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 995\u0026ndash;998. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/Jcm.35.4.995-998.1997\u003c/span\u003e\u003cspan address=\"10.1128/Jcm.35.4.995-998.1997\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson, I. G. Inhibition and facilitation of nucleic acid amplification. \u003cem\u003eAppl. Environ. Microb.\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e, 3741\u0026ndash;3751. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/Aem.63.10.3741-3751.1997\u003c/span\u003e\u003cspan address=\"10.1128/Aem.63.10.3741-3751.1997\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIto, T., Katsushima, H., Tomita, K. M. \u0026amp; Matsumoto, T. Infanticide or predation? Cannibalism by a brown bear in Hokkaido. \u003cem\u003eJapan Ursus\u003c/em\u003e. \u003cb\u003e33\u003c/b\u003e, e13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2192/URSUS-D-22-00006.1\u003c/span\u003e\u003cspan address=\"10.2192/URSUS-D-22-00006.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBellemain, E., Swenson, J. E., Tallmon, O., Brunberg, S. \u0026amp; Taberlet, P. Estimating population size of elusive animals with DNA from hunter-collected feces: Four methods for brown bears. \u003cem\u003eConserv. Biol.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 150\u0026ndash;161. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1523-1739.2005.00549.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1523-1739.2005.00549.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobinson, S. J., Waits, L. P. \u0026amp; Martin, I. D. Estimating abundance of American black bears using DNA-based capture-mark-recapture models. \u003cem\u003eUrsus\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2192/08gr022r.1\u003c/span\u003e\u003cspan address=\"10.2192/08gr022r.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoda, F. et al. Wolf scat detection dog improves wolf genetic monitoring in new French colonized areas. \u003cem\u003eJ. Vertebr Biol.\u003c/em\u003e \u003cb\u003e69\u003c/b\u003e, 20102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.25225/jvb.20102\u003c/span\u003e\u003cspan address=\"10.25225/jvb.20102\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Flamingh, A. et al. Combining methods for non-invasive fecal DNA enables whole genome and metagenomic analyses in wildlife biology. \u003cem\u003eFront. Genet.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 1021004. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fgene.2022.1021004\u003c/span\u003e\u003cspan address=\"10.3389/fgene.2022.1021004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarmacharya, D. B., Thapa, K., Shrestha, R., Dhakal, M. \u0026amp; Janecka, J. E. Noninvasive genetic population survey of snow leopards (\u003cem\u003ePanthera uncia\u003c/em\u003e) in Kangchenjunga conservation area, Shey Phoksundo National Park and surrounding buffer zones of Nepal. \u003cem\u003eBMC Res. Notes\u003c/em\u003e. \u003cb\u003e4\u003c/b\u003e, 516. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/1756-0500-4-516\u003c/span\u003e\u003cspan address=\"10.1186/1756-0500-4-516\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoode, M. J. et al. Capture-recapture of white-tailed deer using DNA from fecal pellet groups. \u003cem\u003eWildl. Biol.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 270\u0026ndash;278. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2981/wlb.00050\u003c/span\u003e\u003cspan address=\"10.2981/wlb.00050\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahman, D. A. et al. The first use of a photogrammetry drone to estimate population abundance and predict age structure of threatened Sumatran elephants. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 21311. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-023-48635-y\u003c/span\u003e\u003cspan address=\"10.1038/s41598-023-48635-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiren, A. P. K. et al. Monitoring animal populations with cameras using open, multistate, N-mixture models. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, e70583. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ece3.70583\u003c/span\u003e\u003cspan address=\"10.1002/ece3.70583\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGaillard, J. M., Festa-Bianchet, M. \u0026amp; Yoccoz, N. G. Population dynamics of large herbivores: variable recruitment with constant adult survival. \u003cem\u003eTrends Ecol. Evol.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 58\u0026ndash;63. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0169-5347(97)01237-8\u003c/span\u003e\u003cspan address=\"10.1016/S0169-5347(97)01237-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoodman, J. P. et al. The ecology of ageing in wild societies: linking age structure and social behaviour. \u003cem\u003ePhilos. T R Soc. B\u003c/em\u003e. \u003cb\u003e379\u003c/b\u003e, 20220464. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1098/rstb.2022.0464\u003c/span\u003e\u003cspan address=\"10.1098/rstb.2022.0464\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Barba, M. et al. DNA metabarcoding multiplexing and validation of data accuracy for diet assessment: application to omnivorous diet. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 306\u0026ndash;323. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1755-0998.12188\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.12188\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJimbo, M. et al. Diet selection and asocial learning: Natal habitat influence on lifelong foraging strategies in solitary large mammals. \u003cem\u003eEcosphere\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, e4105. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ecs2.4105\u003c/span\u003e\u003cspan address=\"10.1002/ecs2.4105\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiofr\u0026iacute;o-Lazo, M., Aurioles-Gamboa, D. \u0026amp; Le Boeuf, B. J. Ontogenetic changes in feeding habits of northern elephant seals revealed by δN and δC analysis of growth layers in teeth. \u003cem\u003eMar. Ecol. Prog Ser.\u003c/em\u003e \u003cb\u003e450\u003c/b\u003e, 229\u0026ndash;241. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3354/meps09562\u003c/span\u003e\u003cspan address=\"10.3354/meps09562\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePolischuk, S. C., Hobson, K. A. \u0026amp; Ramsay, M. A. Use of stable-carbon and -nitrogen isotopes to assess weaning and fasting in female polar bears and their cubs. \u003cem\u003eCan. J. Zool.\u003c/em\u003e \u003cb\u003e79\u003c/b\u003e, 499\u0026ndash;511. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1139/cjz-79-3-499\u003c/span\u003e\u003cspan address=\"10.1139/cjz-79-3-499\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHilderbrand, G. V. et al. The importance of meat, particularly salmon, to body size, population productivity, and conservation of North American brown bears. \u003cem\u003eCan. J. Zool.\u003c/em\u003e \u003cb\u003e77\u003c/b\u003e, 132\u0026ndash;138. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1139/cjz-77-1-132\u003c/span\u003e\u003cspan address=\"10.1139/cjz-77-1-132\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSommer, F. et al. The gut microbiota modulates energy metabolism in the hibernating brown bear. \u003cem\u003eCell. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 1655\u0026ndash;1661. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.celrep.2016.01.026\u003c/span\u003e\u003cspan address=\"10.1016/j.celrep.2016.01.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshikawa, A., Sakamoto, H., Katagiri, S. \u0026amp; Takahashi, Y. Changes in sexual behavior and fecal steroid hormone concentrations during the breeding season in female Hokkaido brown bears (\u003cem\u003eUrsus arctos yesoensis\u003c/em\u003e) under captive condition. \u003cem\u003eJ. Vet. Med. Sci.\u003c/em\u003e \u003cb\u003e65\u003c/b\u003e, 99\u0026ndash;102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1292/jvms.65.99\u003c/span\u003e\u003cspan address=\"10.1292/jvms.65.99\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDalerum, F. et al. Methodological considerations for using fecal glucocorticoid metabolite concentrations as an Indicator of physiological stress in the brown bear (\u003cem\u003eUrsus arctos\u003c/em\u003e). \u003cem\u003ePhysiol. Biochem. Zool.\u003c/em\u003e \u003cb\u003e93\u003c/b\u003e, 227\u0026ndash;234. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1086/708630\u003c/span\u003e\u003cspan address=\"10.1086/708630\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDori, P. et al. Hibernating or not hibernating? Brown bears' response to a mismatch between environmental natural cues and captive management, and its welfare implications. \u003cem\u003ePlos One\u003c/em\u003e. \u003cb\u003e19\u003c/b\u003e, e0306537. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0306537\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0306537\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoriyoshi, M. et al. Patterns of intestinal parasite prevalence in brown bears (\u003cem\u003eUrsus arctos\u003c/em\u003e) revealed by a 3-year survey on the Shiretoko peninsula, Hokkaido, Japan. \u003cem\u003eInt. J. Parasitol-Par\u003c/em\u003e. \u003cb\u003e26\u003c/b\u003e, 101048. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ijppaw.2025.101048\u003c/span\u003e\u003cspan address=\"10.1016/j.ijppaw.2025.101048\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerrero-Garc\u0026iacute;a, G. et al. Non-invasive surveillance of shared pathogens in the Eurasian brown bear (\u003cem\u003eUrsus arctos\u003c/em\u003e) human interface. \u003cem\u003eOne Health-Amsterdam\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e, 100746. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.onehlt.2024.100746\u003c/span\u003e\u003cspan address=\"10.1016/j.onehlt.2024.100746\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShirane, Y. et al. Sex-biased dispersal and inbreeding avoidance in Hokkaido brown bears. \u003cem\u003eJ. Mammal\u003c/em\u003e. \u003cb\u003e100\u003c/b\u003e, 1317\u0026ndash;1326. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/jmammal/gyz097\u003c/span\u003e\u003cspan address=\"10.1093/jmammal/gyz097\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZedrosser, A., Stoen, O. G., S\u0026aelig;bo, S. \u0026amp; Swenson, J. E. Should I stay or should I go? Natal dispersal in the brown bear. \u003cem\u003eAnim. Behav.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 369\u0026ndash;376. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.anbehav.2006.09.015\u003c/span\u003e\u003cspan address=\"10.1016/j.anbehav.2006.09.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBojarska, K. \u0026amp; Selva, N. Spatial patterns in brown bear \u003cem\u003eUrsus arctos\u003c/em\u003e diet: the role of geographical and environmental factors. \u003cem\u003eMammal Rev.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 120\u0026ndash;143. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1365-2907.2011.00192.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2907.2011.00192.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHilderbrand, G. V., Joly, K., Sorum, M. S., Cameron, M. D. \u0026amp; Gustine, D. D. Brown bear (\u003cem\u003eUrsus arctos\u003c/em\u003e) body size, condition, and productivity in the Arctic, 1977\u0026ndash;2016. \u003cem\u003ePolar Biol.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 1125\u0026ndash;1130. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00300-019-02501-8\u003c/span\u003e\u003cspan address=\"10.1007/s00300-019-02501-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026oacute;pez-Alfaro, C., Coogan, S. C. P., Robbins, C. T., Fortin, J. K. \u0026amp; Nielsen, S. E. Assessing nutritional parameters of brown bear diets among ecosystems gives insight into differences among populations. \u003cem\u003ePlos One\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0128088\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0128088\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1. The genomic location of the targets for methylation level analysis and qPCR, and the genes adjacent to them.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdjacent gene\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eFull name of the gene\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eNumber of targets CpG cites\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eDNA positions of target CpG cites in bears\u0026nbsp;\u003cbr\u003e\u0026nbsp;(NCBI sequence ID, position)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eReference\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eDLX5\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDistal-Less Homeobox 5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eNW_026622985.1,\u003cbr\u003e\u0026nbsp;37441371, 37441374, and 37441376\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eTsushima leopard cat\u003csup\u003e30\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eSLC12A5\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSolute Carrier Family 12 Member 5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eNW_020656136.1,\u003cbr\u003e\u0026nbsp;29314689, 29314692, 29314694,\u003cbr\u003e\u0026nbsp;and 29314698\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eBrown bear\u003csup\u003e43\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdjacent gene\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eFull name of the gene\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eqPCR target region\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eReference\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eF2 Intron12\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eCoagulation Factor II (F2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eNW_020656153.1,\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1,273,581\u0026ndash;1,273,681\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003ePolar bear\u003csup\u003e55\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2. Values of mean absolute error (MAE), median absolute error (MedAE), and root mean square error (RMSE) after LOOCV for each model.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" rowspan=\"2\" style=\"width: 30px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" rowspan=\"2\" style=\"width: 35px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eUsed CpGs*\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" colspan=\"3\" style=\"width: 33px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLOOCV model\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMAE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 12px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedAE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" rowspan=\"7\" style=\"width: 30px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSingle regression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.646\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.694\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.843\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.522\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.767\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.575\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e3.100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e2.209\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e4.499\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.645\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e2.346\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.794\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.368\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.477\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.556\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eS-3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.330\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.632\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.414\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.384\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.663\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3.358\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 30px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eElastic net regression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eD-1, -2 S-1, -2, -3, -4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.348\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.519\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3.465\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" rowspan=\"10\" style=\"width: 30px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSupport vector regression\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1, -2, -3, S-1, -2, -3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.163\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.050\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.385\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1, -2, S-1, -2, -3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.123\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e0.984\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.278\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1, -2, S-2, -3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.102\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e0.957\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.267\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eD-2, S-2, -3, -4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.085\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e0.988\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.206\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-2, S-3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.171\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.136\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.220\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1, -2, -3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.493\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.720\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eD-1, -2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.372\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.292\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.550\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-1, -2, -3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.186\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.763\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3.020\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-2, -3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.109\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.921\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.216\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 35px;\"\u003e\n \u003cp\u003eS-3, -4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 9px;\"\u003e\n \u003cp\u003e2.178\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 12px;\"\u003e\n \u003cp\u003e1.169\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11px;\"\u003e\n \u003cp\u003e3.199\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eNote: All values are rounded to the fourth decimal place.\u003c/p\u003e\n\u003cp\u003e*D and S represent DLX5 and SLC12A5, respectively.\u003c/p\u003e\n\u003cp\u003eEach bold value denotes the minimum value for each model.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eTable 3. Coefficients and p-values for GLMM of \u0026Delta;age in the best model (i.e., SVR model, DLX5-2, SLC12A5-2, -3, and -4). The number of host DNA copies used for analysis means host DNA copy number added to the first PCR.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" rowspan=\"2\" style=\"width: 59px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSVR model (DLX5-2, SLC12A5-2, -3, -4)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" rowspan=\"2\" style=\"width: 20px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEstimate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" rowspan=\"2\" style=\"width: 20px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ep-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd height=\"50\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd height=\"36\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 59px;\"\u003e\n \u003cp\u003e\u0026Delta;Age\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" colspan=\"2\" style=\"width: 40px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd height=\"2\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 59px;\"\u003e\n \u003cp\u003e(Intercept)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e1.5067\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.1690\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd height=\"2\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 59px;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026minus;0.7271\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.3120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd height=\"2\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 59px;\"\u003e\n \u003cp\u003eThe number of host DNA copies used for analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026minus;0.3524\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.5540\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd height=\"2\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 59px;\"\u003e\n \u003cp\u003eEnvironment Wild\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026minus;1.7127\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.2310\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd height=\"2\" style=\"width: 0px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;Table 4. Coefficients and p-values for GLMM of |\u0026Delta;age| in the best model (i.e., SVR model, DLX5-2, SLC12A5-2, -3, and -4).\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSVR model (DLX5-2, SLC12A5-2, -3, -4)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 13.5057%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEstimate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11.4798%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ep-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003e|\u0026Delta;Age|\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" colspan=\"2\" valign=\"top\" style=\"width: 24.9855%;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003e(Intercept)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 13.5057%;\"\u003e\n \u003cp\u003e2.3049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11.4798%;\"\u003e\n \u003cp\u003e0.0134\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 13.5057%;\"\u003e\n \u003cp\u003e0.9590\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11.4798%;\"\u003e\n \u003cp\u003e0.1050\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003eThe number of host DNA copies used for analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 13.5057%;\"\u003e\n \u003cp\u003e\u0026minus;0.4609\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11.4798%;\"\u003e\n \u003cp\u003e0.0993\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 71.4451%;\"\u003e\n \u003cp\u003eEnvironment Wild\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 13.5057%;\"\u003e\n \u003cp\u003e\u0026minus;0.1983\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 11.4798%;\"\u003e\n \u003cp\u003e0.8592\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"epigenetic clock, age estimation, brown bear, DNA methylation, feces, wildlife management","lastPublishedDoi":"10.21203/rs.3.rs-9157516/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9157516/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAge is important for understanding the life history of wildlife. We previously established an age estimation method for brown bears based on blood DNA methylation levels. However, collecting blood samples requires invasive procedures. In this study, we established a non-invasive age estimation method based on DNA methylation levels using fecal DNA collected from 43 brown bears of known age living in both captivity and the wild. Bisulfite pyrosequencing was performed to determine the methylation levels of fecal DNA, and the best model was constructed based on four cytosine-phosphate-guanine (CpG) sites: one adjacent to \u003cem\u003eDLX5\u003c/em\u003e and three adjacent to \u003cem\u003eSLC12A5\u003c/em\u003e. The mean absolute error after leave-one-out cross-validation was 2.08 years, and the median absolute error was 0.99 years; these results demonstrate high accuracy. Furthermore, a method was implemented for quantifying host DNA copy numbers, demonstrating that analyses tend to fail or exhibit large estimation errors when the amount of host DNA is insufficient. This is a significant advancement compared to previous techniques that relied on fecal DNA methylation levels. Applying this method to field surveys will greatly contribute to ecological research and the development of appropriate conservation and management strategies for bears, while facilitating future epigenetic clock studies on other animals.\u003c/p\u003e","manuscriptTitle":"Establishment of a non-invasive age estimation method based on fecal DNA methylation levels in brown bears Running title: Fecal DNA-based Epigenetic Clock in Brown Bears","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-01 13:49:43","doi":"10.21203/rs.3.rs-9157516/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-19T12:07:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"112907669431616982462403456666339076407","date":"2026-04-27T06:05:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"112772534354295531427618019654162696314","date":"2026-04-20T20:44:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"154836737970018282962983643531693307779","date":"2026-04-20T17:46:03+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-12T04:37:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"112912473920135574556019361193420875851","date":"2026-04-01T11:25:09+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-30T11:12:04+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-27T14:39:16+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-27T06:26:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-25T08:32:13+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-25T08:24:08+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6de42415-5d72-481a-bc0b-dc6bc74f5db6","owner":[],"postedDate":"April 1st, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-19T12:07:12+00:00","index":73,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":65494528,"name":"Biological sciences/Biological techniques"},{"id":65494529,"name":"Biological sciences/Ecology"},{"id":65494530,"name":"Earth and environmental sciences/Ecology"},{"id":65494531,"name":"Biological sciences/Genetics"},{"id":65494532,"name":"Biological sciences/Molecular biology"},{"id":65494533,"name":"Biological sciences/Zoology"}],"tags":[],"updatedAt":"2026-04-01T13:49:45+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-01 13:49:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9157516","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9157516","identity":"rs-9157516","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}