Roe Deer
Back to the Monitoring Hub
Increase abundance and distribution
Improving
Greater Côa Valley
Portugal
Problem: Carnivores, particularly wolves, are scarce in the Greater Côa Valley. Their return to the landscape is likely to be restricted due to persecution, as well as low availability of prey species, particularly ungulates.
Action: Increasing the abundance of wild prey, particularly roe deer, in the landscape will help reduce the impact of wolves on livestock, reducing conflict with humans and improving attitudes towards the return of this species.
Indicator: The distribution and abundance of wild ungulates (with a focus on roe deer) throughout the Côa Valley indicates prey availability for carnivores. Results will also inform whether translocations may be needed to boost populations of this species.
Methods
The distribution and density of wild ungulate species (roe deer and wild boar) across the landscape were assessed using camera traps and pellet counts. Baseline data were collected between June 2019 and June 2020. Camera-trapping was carried out across six rewilding areas in the Greater Côa Valley, using 35 cameras that were rotated between sites every 28 days. A total of 241 locations were sampled using a systematic grid approach, resulting in a sampling effort of 7,317 camera-trap days across the project area. However, due to theft, malfunction, or problems with the external drive, only records from 230 camera locations could be used in the analysis. The distance between camera locations was based on the minimum home range of roe deer (summer: 409.64 ± 98.20 ha, winter 116.20 ± 17.90 ha; Carvalho et al., 2008). Cameras were sited every 1000-1500 m and placed 30-40 cm above ground to maximise species detectability. Wherever possible, cameras were oriented north to avoid direct sunlight and programmed to take three consecutive photographs, with a recovery time of 30 seconds. Data were also supplemented with camera trapping images from a complementary project.
The camera trap survey was repeated in 2022-2023, in the same areas using the same methodology, to determine any changes in ungulate abundance over the ELSP project period. Thirty cameras were rotated around 180 locations; of these data from 172 locations could be used.
From each camera trap image, the date, time, temperature, species, number of individuals, sex, habitat type, landscape diversity and level of anthropogenic disturbance were recorded. Any photos of the same species from the same camera within 30 minutes of each other were excluded, to reduce the chance of counting the same individual multiple times. The Relative Abundance Index (RAI) was calculated for each species and location as the number of records divided by the number of days the camera was active. RAI estimates were used to map species’ distributions.
Placing a camera trap. Photo: Rewilding Portugal.
Roe deer pellets. Photo: Rewilding Portugal.
To estimate species densities, in 2019-2020 at each sampling location, pellet searches were carried out along a 200m transect (Valente et al. ,2016). Transects were selected to be representative of the main habitat types in the area, and the orientation of each transect was randomly selected, limited by topographic and land use/cover constraints. Transects were surveyed on foot and searches for roe deer pellets were conducted within 1 m either side of the transect line (Marques et al., 2001). Groups with less than 10 pellets were discarded to avoid counting scattered groups multiple times. For each pellet group detected, the following information was recorded: the perpendicular distance from the transect line to the centre of the pellet group, the number of pellets (medium, 10-40 pellets; large, > 40 pellets), the dispersion of pellet groups (aggregated or scattered) and the micro-habitat. In total, 149 pellet groups were recorded across 232 transects and a sampling effort of 46,400 meters. Pellet group data were analysed using the distance sampling method (Buckland et al., 2001) to give an estimate of density. This was not repeated in 2022-2023, and roe deer density was instead estimated from camera trap images, based on the position and speed of each animal detected (ENETWILD-consortium, 2023).
In addition, data from 2019-2020 was used to model environmental favourability for roe deer across the wider region. The model used selected predictor variables related to topographic features, water availability, land cover, landscape structure and anthropogenic pressure. Data from a total of 308 camera trapping locations (128 presences and 180 pseudoabsences) was used for modelling. The study area was divided into a grid composed of 2.256km2 cells, and environmental favourability was modelled for each cell. The final model was selected using forward-backward stepwise selection based on AIC values. This procedure aimed to gather baseline information to decide which areas were suitable for roe deer reintroductions or restocking.
Results
Higher abundance of both roe deer and wild boar was observed in all locations in 2022-23 compared to 2019-20 (Fig. 1), when the ELSP restoration actions got underway. These included two sites, Ermo das Águias and Vale Carapito, that were the focus of rewilding activities. However, these changes have not been tested statistically.
Figure 1: Change in Relative Abundance Index for roe deer and wild boar at the six rewilding locations between the first and second sampling periods. Error bars show standard deviation.
In 2022-2023 the density of roe deer was estimated as between 0.19-6.29 individuals/km², demonstrating the high levels of variability across the landscape.
Results of the environmental favourability modelling exercise are shown in the map below (Fig. 2). Presence of roe deer was significantly positively associated with the percentage of native forests, and roe deer were significantly more likely to occur in valleys with middle or lower slopes. Approximately 70% of the study area presented favourable conditions for roe deer, but only 30% was highly favourable. Full results are available in Rossa et al. (2023).
Figure 2: Environmental favourability map for roe deer across the Côa Valley and wider landscape in 2019-2020. Red areas are least favourable for deer, and green areas are most favourable. Black circles indicate potentially suitable areas for deer reintroduction or restocking. Black arrows indicate potential pathways for spread of existing populations. Hatched lines show areas where the Iberian wolf is also present.
Interpretation
The results of the roe deer monitoring suggest there is no need for roe deer restocking in the study area, since populations appear to be increasing. The density of roe deer population is highly variable; however, roe deer is established, and Rewilding Portugal plan to focus efforts in promoting the westward expansion of this population. These results were used to identify areas that could serve as potential nuclei for roe deer reintroduction, and are strategically located to enhance the natural expansion of existing and established populations. The selected locations also overlap the current wolf range, and in the medium to long term could contribute to the stability of existing wolf packs (Rossa et al., 2023).
Roe deer caught on a camera trap. Photo: Rewilding Portugal.
Wild boar caught on a camera trap. Photo: Rewilding Portugal.
Publications
Rossa M, Torres RT, Linck P, Martins R, Couto A, Aliácar S, Hipólito D, Brotas G, Fonseca C, Carvalho J (2024) Environmental favorability models as a tool to support trophic rewilding initiatives. Restoration Ecology, 32, p.e14048.
Citations
Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L (2001) Introduction to Distance Sampling: Estimating Abundance of Biological Populations. Oxford University Press, Oxford.
Carvalho, P., Nogueira, A., Soares, A. & Fonseca, C. (2008). Ranging behaviour of translocated roe deer in a Mediterranean habitat: seasonal and altitudinal influences on home range size and patterns of range use. Mammalia, 72(2), 89-94.
ENETWILD‐consortium, Guerrasio, T., Pelayo Acevedo, P., Apollonio, M., Arnon, A., Barroqueiro, C., … & Vicente, J. (2023). Wild ungulate density data generated by camera trapping in 37 European areas: first output of the European Observatory of Wildlife (EOW). EFSA Supporting Publications, 20(3), 7892E.
Marques, F. F., Buckland, S. T., Goffin, D., Dixon, C. E., Borchers, D. L., Mayle, B. A., & Peace, A. J. (2001). Estimating deer abundance from line transect surveys of dung: sika deer in southern Scotland. Journal of Applied Ecology, 349-363.
Rossa, M., Torres, R. T., Linck, P., Martins, R., Couto, A., Aliácar, S., … & Carvalho, J. (2024). Environmental favorability models as a tool to support trophic rewilding initiatives. Restoration Ecology, 32(3), e14048.
Valente, A. M., Marques, T. A., Fonseca, C., & Torres, R. T. (2016). A new insight for monitoring ungulates: density surface modelling of roe deer in a Mediterranean habitat. European Journal of Wildlife Research, 62(5), 577-587.
Partners
