Line Østergaard Jensen

Mrs. Line Østergaard Jensen


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 January 2026 in Animals
Verified
Journal Article
Open access
PDF file
JJJohanne Holm Jacobsen
Johanne Holm Jacobsen
Pietro OrlandoPietro Orlando
Pietro Orlando
LJLine Østergaard Jensen
Line Østergaard Jensen
SPSussie Pagh
Sussie Pagh
Cino PertoldiCino Pertoldi
Cino Pertoldi
Rising numbers of wolf (Canis lupus) populations make traditional, resource-intensive methods of wolf monitoring increasingly challenging and often insufficient. This study explores how wolf howls can be used as a new monitoring tool for wolves by applying Artificial Intelligence (AI) methods to detect and classify wolf howls from acoustic recordings, thereby improving the effectiveness of wolf population monitoring. Three AI approaches are evaluated: BirdNET, Yellowstone’s Cry-Wolf project system, and BioLingual. Data were collected using Song Meter SM4 (SM4) audio recorders in a known wolf territory in Klelund Dyrehave, Denmark, and manually validated to establish a ground truth of 260 wolf howls. Results demonstrate that while AI solutions currently do not achieve the complete precision or overall accuracy of expert manual analysis, they offer tremendous efficiency gains, significantly reducing processing time. BirdNET achieved the highest recall at 78.5% (204/260 howls detected), though with a low precision of 0.007 (resulting in 28,773 false positives). BioLingual detected 61.5% of howls (160/260) with 0.005 precision (30,163 false positives), and Cry-Wolf detected 59.6% of howls (155/260) with 0.005 precision (30,099 false positives). Crucially, a combined approach utilizing all three models achieved a 96.2% recall (250/260 howls detected). This suggests that while AI solutions primarily function as powerful human-aided data reduction tools rather than fully autonomous detectors, they represent a valuable, scalable, and non-invasive complement to traditional methods in wolf research and conservation, making large-scale monitoring more feasible.

9 January 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Pietro OrlandoPietro Orlando
Pietro Orlando
LJLine Østergaard Jensen
Line Østergaard Jensen
Cino PertoldiCino Pertoldi
Cino Pertoldi
SPSussie Pagh
Sussie Pagh
Effective wolf monitoring is more critical than ever to support robust population estimates, identify breeding packs, and anticipate and mitigate attacks on livestock. This study evaluates bioacoustic monitoring to estimate wolf population size and detect packs and pups by howling activity, using recordings collected from free-ranging wolves in four different study areas in Denmark and from captive wolves housed in two zoos. It assesses whether howl structure can identify wild individuals and discriminate between current year-pups (aged between four and nine months) and adults. At one wild Location 1, we identified two free-ranging individuals from 40 adult howls by quantifying fundamental-frequency features and applying linear discriminant and multivariate variance analyses; individual classification accuracy was 92%. In captivity, the same workflow yielded 84% accuracy for three wolves at Location 5 and 86% for four wolves at Location 6, including perfect classification for one animal. We examined howls recorded from late August 2021 to February 2022 using maximum fundamental frequency. Across months, mixture modelling and principal component analysis consistently resolved two groups in the wild data, and multivariate tests indicated significant separation each month (p < 0.001), consistent with a pup–adult contrast and the expected autumnal decline in pup frequencies as they mature. A focused analysis restricted to adult-range howls also resolved two groups with very strong multivariate separation (p < 0.001), in line with female–male differences. Overall, passive bioacoustics is an effective, non-invasive approach for wide-area coverage and for inferring pack composition from acoustic evidence alone.

29 April 2026 in Ecologies
Verified
Journal Article
Open access
PDF file
Pietro OrlandoPietro Orlando
Pietro Orlando
LJLine Østergaard Jensen
Line Østergaard Jensen
Cino PertoldiCino Pertoldi
Cino Pertoldi
SPSussie Pagh
Sussie Pagh
Effective wolf monitoring is more critical than ever for supporting robust population estimates, identifying breeding packs, and anticipating and mitigating attacks on livestock. This study evaluates bioacoustic monitoring to estimate wolf population size and detect packs and pups based on howling activity, using recordings collected from free-ranging wolves in four different study areas in Denmark. It assesses whether howl structure can be used to identify wild individuals and discriminate between current-year pups (aged between four and nine months) and adults. At wild Location 1, we identified two free-ranging individuals from 40 adult howls by quantifying fundamental frequency features and applying linear discriminant and multivariate variance analyses; the individual classification accuracy was 92%. We examined howls recorded from late August 2021 to February 2022 using maximum fundamental frequency. Across months, mixture modelling and principal component analysis consistently resolved two groups in the wild data, and multivariate tests indicated significant separation each month (p < 0.001), consistent with a pup–adult contrast and the expected autumnal decline in pup frequencies as they mature. A focused analysis restricted to adult-range howls also resolved two groups with very strong multivariate separation (p < 0.001), but this pattern is considered exploratory and does not constitute confirmed sex discrimination. Overall, passive bioacoustics is an effective, non-invasive approach for wide-area monitoring, individual-level acoustic identification, and detecting the presence of pups within wolf packs; however, inference about pack composition or sex based solely on acoustic clustering requires further validation.

16 December 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
JJJohanne Holm Jacobsen
Johanne Holm Jacobsen
Pietro OrlandoPietro Orlando
Pietro Orlando
LJLine Østergaard Jensen
Line Østergaard Jensen
SPSussie Pagh
Sussie Pagh
Cino PertoldiCino Pertoldi
Cino Pertoldi
Rising numbers of wolf populations make traditional, resource-intensive methods of wolf monitoring increasingly challenging and often insufficient. This study explores how wolf howls can be used as a new monitoring tool fro wolves by applying AI methods to detect and classify wolf howls from acoustic recordings, thereby improving the effectiveness of wolf population monitoring. Three AI approaches are evaluated: BirdNET, Yellowstone's Cry-Wolf project system, and BioLingual. Data were collected using SM4 audio recorders in a known wolf territory in Klelund Dyrehave, Denmark, and manually validated to establish a ground truth of 260 wolf howls. Results demonstrate that while AI solutions currently do not achieve the complete precision or overall accuracy of expert manual analysis, they offer tremendous efficiency gains, significantly reducing processing time. BirdNET achieved the highest recall at 78.5% (204/260 howls detected), though with a low precision of 0.007 (resulting in 28,773 false positives). BioLingual detected 61.5% of howls (160/260) with 0.005 precision (30,163 false positives), and Cry-Wolf detected 59.6% of howls (155/260) with 0.005 precision (30,099 false positives). Crucially, a combined ap-proach utilizing all three models achieved a 96.2% recall (250/260 howls detected). This suggests that while AI solutions primarily function as powerful human-aided data re-duction tools rather than fully autonomous detectors, they represent a valuable, scalable, and non-invasive complement to traditional methods in wolf research and conservation, making large-scale monitoring more feasible.