Prey
Back to the Monitoring Hub
Problem: Predators such as lynx and birds of prey are scarce in this landscape. Low abundance of key prey species, such as rabbit and partridge, is likely to be one barrier to their return. Populations of these prey species are limited by a lack of suitable habitat, as areas of open grassland areas have been lost to shrub encroachment as a result of land abandonment over recent decades. Hunting pressure also keeps prey numbers low. In addition, there is concern that conflict with livestock owners may hamper the recovery of carnivores, if their return leads to an increase in predation on sheep.
Action: Rewilding Portugal has introduced natural grazing by wild horses and tauros, to increase habitat heterogeneity and open up areas of grassland in the landscape and improve habitat for prey species such as rabbit and grey partridge. No-hunting zones are also being established in areas managed by Rewilding Portugal to reduce hunting pressure on key prey species. Boosting the abundance of natural prey will help mitigate the potential impacts of predator recovery on livestock, improving human-wildlife coexistence.
Indicator: An increase in the abundance of rabbit (Oryctolagus cunniculus) and red partridge (Alectoris rufa) in restored areas will indicate greater abundance of key prey species in the food web of this ecosystem, and therefore higher food availability for predators.
Partridge captured on a camera trap in Vale Carapito. Photo: Rewilding Portugal.
Rabbit captured on a camera trap in Vale Carapito. Photo: Rewilding Portugal.
Methods
Red partridge abundance was surveyed using 1 km line transects (Gortázar et al., 2002) located within restoration areas, with natural grazing by Sorraia horses and a hunting ban, and also in paired control areas outside the grazing area and where hunting continued, separated by a minimum distance of 1-2 km. Surveys were carried at two times of year: once before the breeding season (March–April), when the partridge population is at its annual minimum, and again after the breeding season (July–September), coinciding with the population peak. Surveys were carried out in the early morning, within 3 hours of dawn, in good weather conditions. Transects were walked at a constant speed and all partridges seen or heard were recorded, including the number of individuals, age class (adult/juvenile) and time of observation. Each survey was repeated twice within a period of 5 days, to improve the accuracy of the data. Kilometric abundance index (KAI = n/ km) of birds was calculated for each site.
In Vale Carapito, the first site where natural grazing with Sorraia horses was introduced, pre-intervention baseline data were collected in 2020 and a hunting ban was introduced in 2021. In 2021, a second site Ermo das Águias was brought into project management, and baseline data were collected before natural grazing was introduced; a hunting ban was also introduced in 2022. In Vale Carapito, one restoration transect and one control transect were surveyed (Fig. 1). In Ermo das Águias, which is a larger site, four transects were surveyed in the restoration area and two transects at control sites (Fig.1).
Figure 1: Partridge transects in Vale Carapito and Ermo das Águias . The orange polygon delineates the area being restored. Transects are marked by the dotted lines – control in yellow and treatment in white.
Because rabbit densities in the project area are low, an indirect monitoring method of observing latrines was used (Palomares, 2001). Two observers walked at a distance of 2 m each side of a 500 m transect and counted all active rabbit latrines (groups containing more than 20 pellets) observed (LIFE IBERLINCE 2016). A relative abundance index was calculated based on the size and number of latrines detected (Sarmento et al., 2001). Rabbit surveys took place twice a year, in the pre-breeding season (December – January), at the annual population minimum, and in the post-breeding season (June-July), when the population is at its peak. Rabbit surveys took place in the same areas as partridge surveys, with control transects also surveyed in adjacent, non-restoration areas. In Vale Carapito, six transects were monitored (three in restoration areas and three controls, Fig. 2) and in Ermo das Águias, 12 transects were monitored (eight in restoration areas and four controls, Fig. 2).
Figure 2: Rabbit transects in Vale Carapito and Ermo das Águias . The orange polygon delineates the area being restored. Transects are marked by the dotted lines – control in yellow and treatment in white.
In addition, activity of Sorraia horses was monitored to map the impacts of natural grazing across the areas under restoration. Heatmaps were created to visually compare the areas where horses were likely to have an impact with the areas where prey species (rabbits and partridge) were present.
Results
Abundance of the two prey species was low at all sites surveyed over the monitoring period, and this meant statistical analysis of the data was not considered appropriate.
At Ermo das Águias, there was no clear trend in abundance of rabbits either within or outside the rewilding area, although there was some suggestion that numbers have declined between 2021 and 2024 (Fig. 3). Over this period the abundance of rabbits was generally lower in the area with natural grazing compared to the control area outside (overall average of 0.75 rabbits/ha inside the restoration area and 1.15 rabbits/ha outside the restoration area). Rabbit populations are concentrated in rocky outcrops and in high elevation, where they find refuge under the huge granite boulders. During four years of monitoring rabbits have not disperse to lowland areas where Sorraia horses spent most of their time, therefore overlap between both species is restricted to a few areas that horses use with low intensity (Fig. 5).
In Vale Carapito the abundance of rabbits showed greater fluctuations, with higher overall average abundance in the grazed area (0.75 rabbit/ha) than outside in the control area (0 rabbits/ha) (Fig. 3). Rabbits in Vale Carapito are only present in a part of the area, and in 2024 some latrines were close to a high use area by horses (Fig. 5).
Figure 3: Rabbit abundance measured in summer and winter in the two areas, across both restored and control (unrestored) sites.
Similarly, there was no clear trend in partridge abundance at Ermo das Águias over four years of the ELSP project (Figure 4). Over all visits, the average kilometric abundance index for partridge was 2.2 within the restoration area and 2.4 outside the restoration area. Partridge distribution overlaps with horses’ and has remained stable over time (Fig. 5). No changes in territories or other spatial patterns were observed due to horse activity.
Partridge numbers in Vale Carapito were also low both within and outside the restoration area (Figure 4), with an average kilometric abundance index of 2.6 and 0.1 respectively. Partridges were well distributed throughout Vale Carapito, with a slight concentration of points near one of the horse activity hotspots, most of which were observed opportunistically because the transect passes nearby but does not overlap the hot spot (Fig. 5).
Figure 4: Partridge abundance measured in summer and winter in the two areas, across both restored and control (unrestored) sites.
Figure 5: Heat maps showing Sorraia horse activity (red, with darker shades representing more time spent in the area) overlap with areas of rabbit and partridge presence (transects in turquoise, and detected presence in each year in coloured circles).
Interpretation
Rewilding Portugal aims to increase prey abundance in the landscape by using natural grazing to create a more open and diverse mosaic habitat, and implementing hunting bans. However, the low density of rabbits in the landscape and their fluctuating trends do not yet allow a robust assessment of the effects of rewilding on their populations. Many other factors can affect the abundance of rabbits, including disease, weather, predation and food availability. During the four years of this study, it appears that the influence of the other factors may be stronger than any effects of restoration activities on rabbit population. Similarly, so far there are no clear trends in partridge numbers linked to the activities of Rewilding Portugal; although there may be some suggestion of a positive trend in partridge numbers inside the restoration area at Vale Carapito, this needs to be confirmed by monitoring in the longer term.
The relative abundance index (Sarmento et. al, 2004) that was used to assess rabbit abundance gives weighting to the size of the latrine, rather than just the presence or absence of latrines. Larger populations of rabbits tend to make disproportionately larger latrines compared to smaller populations. Initial results suggest that all areas surveyed for this indicator have rabbit populations capable of sustaining vagrant Iberian Lynx.
Rabbit and partridge populations in the landscape will continue to be monitored following the same approach under the LIFE LUPI LYNX project. This will improve understanding of the factors affecting partridge and rabbit populations and any effects of rewilding actions in the longer term.
Lessons learned
In the first monitoring season, an established pellet count method was used to assess rabbit densities (Fernandez-de-Simon et al., 2011). However, due to the very low densities of rabbit in the study areas, more than 99% surveys found zero pellets, making it impossible to estimate abundance; the latrine count method was therefore used in subsequent sampling sessions. Although this latrine count approach is recommended for areas with low densities of rabbits, it was found to have limitations in areas of very low densities such as the Greater Côa Valley. For example, changes were visually observed in the use of different areas by rabbits in Vale Carapito. At low densities rabbits tended to cluster spatially in groups of a number of families; sometimes pellets were grouped close to the transects and so they were detected in the counts and other times pellets were just outside the survey area, and therefore given a zero count. To try to reduce these biases, a different approach was tested, in which circular buffers were monitored in addition to the transects, to improve understanding of the distribution and movement of rabbit colonies. 50 m radius circles were placed either at the centre of a large latrine or at random in the landscape. Although counts from these buffers contained many more latrines than the transects alone, they were much more time consuming to survey and did not cover such a large area. However, if sufficient capacity is available, counting latrines both along transects and within buffer areas should yield more robust results where rabbit densities are very low.
Citations
Fernandez-de-Simon, J., Díaz-Ruiz, F., Cirilli, F., Tortosa, F. S., Villafuerte, R., Delibes-Mateos, M., & Ferreras, P. (2011). Towards a standardized index of European rabbit abundance in Iberian Mediterranean habitats. European Journal of Wildlife Research, 57(5), 1091-1100.
Gortázar, C., Villafuerte, R., Escudero, M. A., & Marco, J. (2002). Post-breeding densities of the red-legged partridge (Alectoris rufa) in agrosystems: a large-scale study in Aragón, northeastern Spain. Zeitschrift für Jagdwissenschaft, 48(2), 94-101.
Palomares, F. (2001). Comparison of 3 methods to estimate rabbit abundance in a Mediterranean environment. Wildlife society bulletin, 578-585.
Sarmento, P, Cruz, J., Tarroso, P. & Gonçalves, P. (2001). Recuperação do habitat e presas do lince-ibérico na Serra da Malcata. Project Life Habitats. 2º Relatório de Progresso. ICN/RNSM. Unpublished internal report.
Sarmento, P., Cruz, J., Monterroso, P., Tarroso, P., Negrões, N. & Ferreira, C. (2004). The Iberian lynx in Portugal. Status survey and conservation action plan. Instituto da Conservação da Natureza (ICN).
Partners
