Every spring, countless volunteers, farmers, and hunters are involved in searching for roe deer fawns on agricultural land before mowing starts. Due to the high forage content of meadows and field forage areas, roe deer females prefer these areas during fawning season, especially as they offer sufficient cover for the young from natural predators. However, the natural instinct of the offspring to hide in the face of danger in the first few weeks of life can have dangerous consequences - namely when this behaviour coincides in space and time with spring mowing activities by humans.
Roe deer - widespread habitat generalists
Fig. 1: To determine the age, the project team measures the fawn and secures it in a box during mowing. As soon as the mowing is finished, it is released again. Photo: Alissa Lüpke
Roe deer is the most widespread ungulate species in Europe and uses a wide range of habitats. Originally, roe deer was considered as a species adapted to forests (Putman 1988), but they also use various landscapes; diverse and structurally rich areas in particular (Said et al. 2005, Morellet et al. 2011). The animals adapt their small-scale habitat selection to their specific needs and conditions at the respective time of year. In spring, fields and meadows provide abundant food resources for the females, as well as sufficient cover for their offspring. Roe deer fawns are hiders, meaning that they left hidden by their mothers in the first weeks of life. If they perceive danger, they remain motionless in hiding. However, their supposedly safe bed site, in combination with their hiding strategy, can entail a high risk if human land use, such as the mowing of grassland, coincides in space and time.
Choice of bed sites at two spatial scales
As part of the research project “Wildlife Rescue Strategies”, we accompanied drone pilots from all over Bavaria in their search for fawns over a period of three years. In addition, via a citizen science platform, volunteers were also able to report their bed sites independently via the Bavarian Wildlife Portal and thus actively participate in the project. During this process, data were collected on the characteristics of the preferred bed sites, as well as on the characteristics of meadows where no fawns were found. Our aim in collecting the data was to gain deep knowledge of the exact locations where the fawns were hiding. Additionally, we wanted to separately determine the factors that the female and the fawn prefer when choosing a bed site. Although the exact choice of the bed site is primarily made by the fawn itself (Bubenik 1967), its decision is not entirely independent, as the mother has already pre-selected a certain area within her home range as a place to give birth and raise the offspring. Because of the low movement rates of the roe deer fawns in the first days of life, the choice of bed sites is limited by this maternal “pre-selection”.
What influence do mother and offspring have?
More than 600 fawn bed sites and more than 450 fields where no fawns were found were included in this comprehensive study in Bavaria. We analysed the data collected at two spatial levels. This approach made it possible to estimate the influence of the female on the approximate choice of the site (landscape level) and to investigate the exact choice of bed site by the fawn within this “pre-selection” (field level) (Figure 2). We used geo-information data to characterise the areas within a radius of 100 m around the bed site at the landscape level. Here, factors such as the structural richness of the landscape, the presence of human settlements, distance to roads and forests, the number of mown fields in the surrounding of the bed site at which the fawn was found, and the NDVI (Normalised Difference Vegetation Index, as an indicator of food availability for the doe) were included and compared statistically with the areas where no fawns were found. The statistical approach used (GAMLSS) made it possible not only to compare the fields with and without fawns but also to investigate which factors favoured the number of fawns hidden.
Fig. 2: Visualization of the two spatial levels analysed for assessing the preferred bed sites of fawns. The landscape level reflects the female’s choice of the parturition and rearing site, within which the fawn selects its exact bed site at the field level.
At the smaller spatial level (field level), our recording teams took detailed measurements at the exact bed site and a random reference area 50 meters away. Parameters such as the grass height or the distance to the nearest woody structure were recorded here. The visibility of the fawns for potential predators (in our case the red fox) was also assessed. For this purpose, a fawn dummy was placed in the original bed site, and the minimum distance from which a potential predator could see the fawn was measured. Furthermore, the age of the fawn found was estimated in order to take into account any age dependency. The data was analysed using a logistic regression analysis, in which a bed site was compared with its corresponding reference area (clogit model). This made it possible to analyse which factors determine the exact choice of a bed site at a very small spatial level (within the field).
Not all grassland is the same
Fig. 3: Results of the bed site analysis for both spatial levels. The statistically significant variables are shown in black. An upward arrow means under a) an increasing probability of finding a fawn correlates with an increase in the given variable; under b) an increasing number of fawns found with an increase in the given variable; and under c) a selection for longer distances with an increase in the variable.
The results obtained suggest that there were certainly differences between the landscape and field levels, but also similarities. For example, the females (landscape level) tended to select diverse areas that were not in the immediate vicinity of human settlements to leave their fawns hidden. Proximity to roads had no negative influence per se on the probability of fawns found and sites closer to woodland or hedgerow structures were preferred.
The number of fawns being found in an area was higher if there were woody structures (hedges or forest) in the immediate proximity. The number of surrounding fields that had already been mown also had a major influence: More fawns were found whenever only a few fields in the vicinity still offered good hiding possibilities. Hence it can be concluded that the fawns’ need for cover has a decisive influence on the choice of the bed site at the landscape level. This is consistent with the results of Linnell et al. (2004). In this study from Norway, the scientists showed that the selection of agricultural land for hiding sites was determined by grass growth and the harvest of the arable land.
At the field level, the results show that, compared to a random comparison site, the fawns’ bed sites tended to be in areas with dense vegetation, i.e. where it is harder for potential predators to see them. The height of the grass or crop alone, on the other hand, played only a subordinate role. At this level, fawns favoured bed sites that were not in close proximity to the forest-field edge or a road. The fawns thus did not hide directly on the edges, but further inside the fields. Hence, the selection of the exact bed site (field level) is in contrast to the selection at the landscape level. The females favoured short distances to forests or hedges, which offered them cover and feeding possibilities. We interpret this as an aspect of their predator avoidance strategy. As foxes prefer to orientate along structures such as hedges, a certain distance from these may be advantageous for the fawns. However, these structures not only offer protection for the fox, but also for the female, giving it the opportunity to intervene and defend her fawn in the case of danger (Jarnemo, 2004). Contrary to our assumptions, no significant differences were found between different age groups. There was also no evidence of an influence of the topography on the selection of bed sites (Figure 3).
Based on solid and objective data, the results obtained should help us to understand the factors influencing the selection of bed sites and contribute to a better assessment of the landscape in which the females give birth. Taking regional information and conditions into account, the data can help those responsible to plan a wildlife-friendly mowing regime.
Fig. 4: Fawn found on an area being mown. Photo: Alissa Lüpke
Summary
The results of this study highlight that grassland is a frequently used habitat for roe deer fawns to hide. The natural instinct of fawns to hide in dense vegetation to protect themselves from predators also increases the risk of being overlooked during spring mowing. This is particularly challenging because it is difficult to spot a well-hidden fawn, especially in these dense grass stands. This underlines the importance of technical methods such as drones for saving fawns. The meadows should be searched particularly carefully, especially in richly structured landscapes and when many fields in the area have already been mown. Details of the original analysis can be found in Baur & Kauffert et al. (2023).
The project “Reduction of Mowing-Related Mortality in Wildlife Based on the Example of Roe Deer Fawns” was funded by the Bavarian State Ministry of Food, Agriculture, Forestry and Tourism (period: 01.01.2020 to 30.06.2024) and carried out in cooperation with the Bavarian State Institute of Agriculture, Institute of Agricultural Engineering and Animal Husbandry and the Technical University of Munich, Professorship of Ecoclimatology and the Working Group for Wildlife Biology and Wildlife Management.