Thomas Vogler

• Vogler, Thomas; Devineau, Olivier; Lindgren, Finn Kristian; Strindberg, Samantha; Ganbaatar, Oyunsaikhan & Altansukh, Nanjid [Show all 7 contributors for this article] (2026). Comparing distance sampling and spatial point process models for estimating wide-ranging, aggregated ungulate populations. Global Ecology and Conservation. 69. doi: 10.1016/j.gecco.2026.e04295. Full text in Research Archive Show summary

  Ungulate populations are declining globally due to increasing anthropogenic pressures on habitats and migration routes, highlighting the importance of protected areas for their conservation. The protected areas in the Mongolian Gobi provide a critical stronghold for several threatened ungulate species, including the khulan (Asiatic wild ass, Equus hemionus) and goitered gazelle (Gazella subgutturosa). Although long-term ungulate population monitoring is essential to assess the effectiveness of conservation measures, it is challenging due to low population densities, fission-fusion social dynamics, and highly variable environmental conditions. We analysed data from large-scale ground surveys conducted in 2010, 2015, and 2022 in the Great Gobi B Strictly Protected Area (GGB), comparing two analytical approaches: design-based point transect distance sampling and Bayesian point process modelling. The latter produces spatially explicit density surfaces that account for spatial autocorrelation alongside abundance estimates. Both methods revealed consistent population trends: khulan numbers declined over the study period, while the gazelle population increased threefold. The Bayesian approach provided greater precision and identified aggregation hotspots. These findings underscore the benefits of modelling the spatial structures within a point process framework for aggregated species in landscapes with uneven and temporally variable resource distributions. We recommend continuing the point transect surveys at least every 5 years, ideally expanding the coverage to suitable habitats adjacent to the protected area. In addition, ranger observations on distribution, group size, reproduction and mortality are needed to help identify drivers of population change and to obtain annual minimum population counts.

• Vogler, Thomas; Thorsen, Neri Horntvedt & Pedersen, Simen (2025). Saving Bambi from the mower? Using a drone with thermal camera to evaluate a low‐tech scaring technique to reduce roe deer fawn mortality during grass harvest. Wildlife Biology. ISSN 0909-6396. doi: 10.1002/wlb3.01450. Full text in Research Archive Show summary

  Roe deer is a species that hides their neonates as an anti‐predator strategy. This may prove efficient against mammalian predators, such as the red fox; however, it might be an ecological trap as large numbers of fawns are killed by tractors with harvesters each year during grass harvest. Here, we evaluate a low‐tech, low‐cost method, using scaring flags placed in the field the evening before grass harvest, intended to stimulate the roe deer doe to remove any fawns from the field. We evaluate the effects of the scaring flags by counting the number of fawns in the fields using a multi‐rotor drone equipped with a thermal camera. The drone proved to be an efficient tool for detecting roe deer fawns. However, contrary to other studies, we found limited or no effect of placing the scaring flags in the field. This discrepancy highlights the fallacy of generalizing based on single case studies. There is still a need to develop low‐cost, easily applicable tools to reduce roe deer fawn mortality during grass harvest. However, with the increasing technological development, drones with thermal cameras are a promising tool for wildlife management and monitoring in the future.

• Østbye, Kjartan; Hassve, Marius Hagen; Tamayo, Ana-Maria Peris; Hagenlund, Mari; Vogler, Thomas & Præbel, Kim (2020). “And if you gaze long into an abyss, the abyss gazes also into thee”: four morphs of Arctic charr adapting to a depth gradient in Lake Tinnsjøen. Evolutionary Applications. 13(6), p. 1240–1261. doi: 10.1111/eva.12983. Full text in Research Archive Show summary

  The origin of species is a central topic in biology. Ecological speciation might be a driver in adaptive radiation, providing a framework for understanding mechanisms, level, and rate of diversification. The Arctic charr <i>Salvelinus alpinus</i> L. is a polymorphic species with huge morphological and life‐history diversity in Holarctic water systems. We studied adaptive radiation of Arctic charr in the 460‐m‐deep Lake Tinnsjøen to (a) document eco‐morphology and life‐history traits of morphs, (b) estimate reproductive isolation of morphs, and (c) illuminate Holarctic phylogeography and lineages colonizing Lake Tinnsjøen. We compared Lake Tinnsjøen with four Norwegian outgroup populations. Four field‐assigned morphs were identified in Lake Tinnsjøen: the planktivore morph in all habitats except deep profundal, the dwarf morph in shallow‐moderate profundal, the piscivore morph mainly in shallow‐moderate profundal, and a new undescribed abyssal morph in the deep profundal. Morphs displayed extensive life‐history variation in age and size. A moderate‐to‐high concordance was observed among morphs and four genetic clusters from microsatellites. mtDNA suggested two minor endemic clades in Lake Tinnsjøen originating from one widespread colonizing clade in the Holarctic. All morphs were genetically differentiated at microsatellites (<i>F</i> _ST: 0.12–0.20), associated with different mtDNA clade frequencies. Analyses of outgroup lakes implied colonization from a river below Lake Tinnsjøen. Our findings suggest postglacial adaptive radiation of one colonizing mtDNA lineage with niche specialization along a depth–temperature–productivity–pressure gradient. Concordance between reproductive isolation and habitats of morphs implies ecological speciation as a mechanism. Particularly novel is the extensive morph diversification with depth into the often unexplored deepwater profundal habitat, suggesting we may have systematically underestimated biodiversity in lakes. In a biological conservation framework, it is imperative to protect endemic below‐species‐level biodiversity, particularly so since within‐species variation comprises an extremely important component of the generally low total biodiversity observed in the northern freshwater systems.

View all works in NVA

• Vogler, Thomas; Jahren, Torfinn; Mayer, Martin & Kvasnes, Mikkel Andreas Jørnsøn (2025). Monitoring capercaillie and their lek sites using a drone mounted thermal camera. doi: 10.6084/M9.FIGSHARE.30206878.V1. Full text in Research Archive Show summary

  We investigated the use of drones equipped with thermal cameras to monitor capercaillie. Grid flights were conducted over the last known GPS positions of tagged birds. Thermal images were manually reviewed to register detections. Following the grid flights we carried out approach flights by lowering the drone from 60 m to 20 m in 10-m increments, hovering for one minute at each step to record behavioural responses. Our results indicate that capercaillie did not exhibit flight responses to the drone. However, increased vigilance was observed at very low altitudes, typically expressed as repeated upward glances while continuing to forage. During active lekking activity, no response was registered. Detection probability of GPS-tagged birds was 40%. There was no snow cover during the study period and we expect detection rates to improve under snow covered conditions, which is typical for lekking season.

• Vogler, Thomas; Kaczensky, Petra; Devineau, Olivier; Payne, John & Salemgareyev, Albert (2025). Monitoring plains ungulates using drone images and computer vision. Full text in Research Archive Show summary

  Monitoring wildlife populations is essential for effective conservation, yet obtaining accurate population estimates remains challenging. Species differ in their habitat use, behaviour, and spatial distribution, requiring survey methods tailored to their ecology. The Asiatic wild ass (Equus hemionus) and goitered gazelles (Gazella subgutturosa) inhabit open landscapes of Central Asia, occur at low densities and form variable group sizes with uneven spatial distributions – characteristics that make them difficult to monitor. Aerial surveys are well suited to address these challenges. Recent advances in drone technology have made it possible to replace manned aircraft with drones, offering more cost-effective and logistically flexible survey options. In June 2024, we conducted a large-scale drone survey in Altyn Emel National Park, Kazakhstan. Using a fixed wing drone equipped with dual angled cameras, timed to capture images every three seconds, we completed eight strip transects spaced 1 km apart, covering a total flight distance of over 1,500 km. The dual-camera setup nearly doubled the photographed strip width and produced over 40,000 images. To enable automated detection, we annotated images from previous surveys containing over 3,000 observations of Asiatic wild ass and almost 4,000 of goitered gazelle. These annotations were used to train an object detection model, which achieved strong performance (recall = 0.86, precision = 0.92) in detecting and classifying individuals. We expect similarly high detection and classification performance in the 2024 dataset, as the training images include substantial variation in habitat, lighting, image quality, and ground resolution. However, due to the lack of camera stabilization, many images are blurred, making some animals difficult to detect even during manual annotation. We therefore assume that some individuals remain undetected, both manually and by the model, suggesting that our final population estimates may be conservative. Our results demonstrate that, with a well-trained computer vision model, future surveys can efficiently estimate populations of these species and support informed conservation decisions. However, since most of the training data originated from Altyn Emel, model performance may not generalize to other regions with different environmental backgrounds. Additional training data from diverse sites will be required to extend the model’s broader applicability.

• Zimmermann, Barbara & Vogler, Thomas (2021). Drone field demonstrations in Norway. Full text in Research Archive
• Pedersen, Simen & Vogler, Thomas (2017). Forsøk i Nærøy skal finne løsninger mot rådyrdød i slottonna. [Journal]. Namdalsavisa. Full text in Research Archive
• Zimmermann, Barbara; Tofastrud, Morten & Vogler, Thomas (2017). Storfe på skogsbeite - Sommeren 2016. Full text in Research Archive
• Zimmermann, Barbara; Tofastrud, Morten; Vogler, Thomas; Bakmann, Ole & Skogholm, Bjørn (2016). Cattle in the woods: Sustainable beef cattle production on forest pastures. Full text in Research Archive
• Vogler, Thomas; Altansukh, N.; Ganbaatar, Oyunsaikhan; Sukhbaatar, Dalai; Devineau, Olivier & Kaczensky, Petra (2023). Steppe ungulate count in Great Gobi B Strictly Protected area 2022. Universitetet i Innlandet. ISSN 9788283804003. Full text in Research Archive Show summary

  The plains of Great Gobi B Strictly Protected Area (subsequently “Great Gobi B”) in southwestern Mongolia are home to three endangered wild ungulates, khulan (Equus hemionus), takhi (Equus ferus przewalskii), and goitered gazelle (Gazella subgutturosa). With Mongolia holding the largest populations of these species, their conservation is of global importance. To assess the effectiveness of conservation efforts, robust survey methods are needed to monitor population development. In late summer 2022 we conducted the 3rd plains ungulate count in Great Gobi B to estimate population size of khulan and goitered gazelle and describe population development since the counts in 2010 and 2015 and tested the method for estimating the growing population of reintroduced Przewalski’s horses. We conducted Distance Sampling point counts at 101 observation points over 6 counting events (at 19:00, 7:00, 9:00, 11:00, 13:00, and 15:00) at each observation point. During each counting event the observers scanned the surroundings using binoculars and registered species, number of animals, time, bearing and distance for each observation. Over the 606 counts, we observed 5,744 khulan, 3,150 gazelles and 922 Przewalski’s horses. Using these observations, we created global models (over all observation points and counting events) using the distance analysis framework and selected models with the best fit based on AIC values. Populations estimates for 2022 were 5,204 (95% CI = 2,121 – 12,771) khulan, 10,980 (95% CI = 7,473 – 16,132) goitered gazelles, and 1,288 (95% CI = 213 – 7,776) Przewalski’s horses within the 13,000 km2 survey area. Population estimates of both khulan and goitered gazelle suggested an increase from 2010 to 2015, while the 2022 estimated is closer to the 2010 estimates for khulan and in between for goitered gazelles. However, confidence intervals, especially for khulan, are large and population development cannot be determined conclusively. The uncertainty in the khulan estimate for 2022 was caused by the combination of a highly clumped distribution due to draught conditions and a large variation in group sizes. Goitered gazelles were more evenly distributed and group sizes less variable. Estimating Przewalski’s horse population resulted in a gross overestimate because of the knowledge rangers had about the location of Przewalski’s horse groups from intensive weekly monitoring. The data suggests that they specifically looked for groups, as Przewalski’s horses were detected almost equally likely over all distance categories. Confidence intervals were large because the population is still small (numbering just over 400) and distribution was also highly clumped due to the drought condition. The 2022 population estimates for khulan and goitered gazelles and the comparison with the estimates from 2010 and 2015 should be regarded as preliminary because: 1) The 3 surveys were analysed using slightly different distance analysis frameworks and 2) Distance Sampling does not take the spatial distribution of groups into account and we plan to explore such methods before we will reanalyse the 3 surveys within the same analysis framework.

• Kjønsberg, Marius; Pedersen, Simen; Vogler, Thomas & Johnsen, Kaja (2022). Kurs for taksering av hønsefugl. Linjetaksering med hund – Distance sampling. Universitetet i Innlandet. ISSN 9788283803167. Full text in Research Archive Show summary

  Hønsefuglportalen er en webportal for registrering og håndtering av data fra taksering av hønsefugl i Norge. I forbindelse med Hønsefuglportalen har Høgskolen i Innlandet (INN), i samarbeid med Norsk Institutt for Naturforskning (NINA), utviklet et kurs i taksering av hønsefugl basert på linjetaksering med avstandsmetoden (distance sampling). Kurset er utviklet for feltpersonell og har som hensikt å øke kvaliteten på dataene til Hønsefuglportalen, som bidrar til mer nøyaktige tetthetsestimater. Dette kursheftet danner det teoretiske grunnlaget for takseringskurset, og er bygd opp av fire hovedtemaer: bakgrunn for kurset, overvåking av hønsefugl, distance sampling som metode, og linjetaksering med hund.

• Mathisen, Karen Marie & Vogler, Thomas (2020). Hva gjør vi når det blir bråstopp i fôring av hjortevilt? - Erfaringer fra Stor-Elvdal. Universitetet i Innlandet. ISSN 9788283802382. Full text in Research Archive Show summary

  Vinterfôring av hjortevilt er en vanlig forvaltningspraksis i Norge. I Stor‐Elvdal kommune har de brukt silofôring av elg om vinteren i over 30 år, for å holde dem unna vei og jernbane. Formålet er å redusere antall påkjørsler, i et område som er svært utsatt for viltulykker. Da skrantesjuke (CWD) ble oppdaget i Norge i 2016, ble det påfølgende år innført et forbud mot fôringsplasser for hjortevilt, for å unngå konsentrasjoner av dyr og økt smittefare. Dermed ble et viktig verktøy i lokal forvaltning tatt bort. Denne rapporten undersøker hva som skjedde vinteren 2018/2019 i Stor‐Elvdal, når det ble bråstopp i vinterfôring på grunn av silomangel, og diskuterer dette i lys av det nye forbudet mot vinterfôring, hva er alternativene? Vi har også undersøkt problemet med vinterhogster langs riksveier, og om det lønner seg å frakte hogstavfall vekk fra veien og til etablerte fôringsplasser. Vi fant at frakt av hogstavfall med traktor og tilhenger, var svært dyrt sammenlignet med silofôring av elg. Dette skyldes mest fraktkostnadene, blant annet er fôrtettheten av silo mye høyere enn i furubar, og en transport kan frakte mye mer mat. På det annen side er furubar gratis, og er en del av elgens naturlige vinterdiett. Heltrefelling i landskapet var billigere enn frakting av hogstavfall. Elgens utnyttelse av hogstavfall på fôringsplasser var svært høy sammenlignet med tidligere studier av bruk av hogstavfall i landskapet, fra 90‐100%. Dette kan skyldes at mange elg oppsøker tidligere fôringsplasser. Beitetrykket i landskapet økte noe vinteren 2019 sammenlignet med året før, men påkjørsler var høyere året før enn året uten fôring. Dette skyldes antakelig at snøforhold og vinterlengde har mye sterkere effekt på påkjørsler enn det fôring har, og vinteren 2018/2019 var relativt mild med lite snø. På kort sikt, har kommunen fått dispensasjon til å fortsette å fôre. Men på lang sikt, er det nødvendig å tenke på alternativer. Vi foreslår å øke naturlig fôrtilgang gjennom skogskjøtsel og avvirkning på en viss avstand fra vei og jernbane, og minske fôrtilgangen i sonen nærme vei og jernbane. Her kan man bygge opp et fôrrikt landskap over tid, med ulike tiltak. På vei og jernbane, kan man prioritere fleksible tiltak i perioder med høy risiko for påkjørsler, og her bør man prøve ut tiltak som tar i bruk ny teknologi.

• Vogler, Thomas & Kjønsberg, Marius (2017). Elgbeitetaksering Finnemarka 2017. Høgskolen i Innlandet. Full text in Research Archive

View all works in NVA