Introduction
Semi-natural open areas such as grasslands, heathlands or wetlands often act as important refuges for rare, highly specialized plant and animal species (Luoto et al.
2003; Warren et al.
2010; Benthien et al.
2018; Riesch et al.
2020) as a result of natural open areas being degraded or lost to land-use change (Carbutt et al.
2017). These ecosystems and their associated species, however, are increasingly under pressure due to the natural succession towards closed forests (Pakeman et al.
2003; Buchholz et al.
2013; Koch et al.
2015). Open areas with a history of anthropogenic use (e.g., military training sites or extensive pastures) particularly experience encroachment by shrubs or trees often resulting in a loss of biodiversity after human activities cease or decrease (Luoto et al.
2003; Warren et al.
2010).
In the past decades, targeted low-intensity grazing with semi-feral or domesticated livestock—also called conservation grazing (sensu Bailey et al.
2019)—has gained popularity as a tool for maintaining or restoring semi-natural open landscapes by counteracting natural succession (e.g., van Wieren
1995; Dostálek and Frantík
2008; Jauregui et al.
2009). Through mechanisms such as browsing, trampling, defecation, and seed dispersal, conservation grazing can additionally increase plant species diversity, structural diversity, and species turnover (Bakker et al.
1983; Dostálek and Frantík
2008; Benthien et al.
2018; Riesch et al.
2020). Moreover, low-intensity grazing regimes have also been found to enhance overall faunistic biodiversity (van Wieren and Bakker
2008; but see Reading and Jofré,
2015). For example, beneficial effects have been described for species richness and density of birds (Zalba and Cozzani
2004) or species richness and turnover of spiders (Dennis et al.
2015). Further, low-intensity grazing can help to maintain unique species assemblages of ecosystems sensitive to shrub- or tree-encroachment, for instance in carabid beetles (Schirmel et al.
2015). Therefore, conservation grazing has become an important management tool in nature conservation when it comes to maintaining or restoring open landscapes such as grass- and heathland (Newton et al.
2009). However, it is not currently clear to what extent semi-feral or domesticated herbivores interact with and influence wild ungulate herbivores, as research investigating the ecological consequences of conservation grazing has mainly focused on impacts on vegetation (Gallet and Roze
2001; Jauregui et al.
2009; Benthien et al.
2018) and smaller animal species (van Wieren and Bakker
2008; Schirmel et al.
2015; Schwerk et al.
2021). But also wild herbivores have the potential of maintaining open landscapes and enhance biodiversity (van Wieren and Bakker
2008). Red deer (
Cervus elaphus) are large ungulates that display grazing-type feeding behavior and often utilize open areas such as grass- and heathland for foraging (Wolff and Horn
2003; Godvik et al.
2009; Meißner et al.
2012). Riesch et al. (
2019) found in a field experiment that the quantity of biomass removed by wild red deer in semi-natural grass- and heathland is comparable to that theoretically achieved by livestock at stocking rates recommended for conservation grazing. Furthermore, a study by Hester and Baillie (
1998) on enclosed heathland plots showed that the grazing impact of red deer can even exceed that of sheep, even when red deer were present at lower densities than sheep (12 vs. 8 animals h
−1). Moreover, Riesch et al. (
2020) observed an increase in vegetation height and the encroachment of woody vegetation following the exclusion of red deer. In the case of conservation grazing the similar use of resources might lead to competition between domesticated and wild herbivores such as red deer.
In the anthropogenic landscapes of central Europe, large semi-natural open areas nestling within forested areas often represent important refuges for wildlife such as red deer, which typically use these tracts of land for foraging (van Wieren and Bakker
2008) and, to a lesser (but equally important) extent, for mating (Meißner et al.
2012). These areas are usually less exposed to human activities (e.g., agriculture, recreational activities, traffic) and might enable red deer to adopt less disturbed activity patterns: Red deer are often described to have crepuscular activity rhythms showing peaks of activity around sunrise and sunset (e.g., Clutton-Brock et al.
1982; Godvik et al.
2009; Ensing et al.
2014). However, in absence of disturbance, these peaks seem to be less pronounced (Kamler et al.,
2007), and diurnal activity increases (Ensing et al.
2014). Red deer are known to be sensitive to disturbances (Edge and Marcum
1985; Czech
1991; Sibbald et al.
2011), especially in open areas, as these lack potential cover in which to hide (Jayakody et al.
2008; Stankowich
2008). They are also reportedly sensitive to rapid movements and noise, either of which usually induces a flight response (Frid and Dill
2002), and it was found that disturbances can still affect their spatial behavior at large distances away from the actual source of disturbance (Edge and Marcum
1985). Several studies have reported that free-ranging red deer avoid areas used by cattle (Stewart et al.
2002; Coe et al.
2004; Pruvot et al.
2014), whilst a study by Hester et al. (
1999) found that red deer and sheep are weakly affected by each other’s presence or absence in the extensive heath moorland of north Scotland. Nevertheless, to the best of our knowledge, no research has investigated how free-ranging red deer react to conservation grazing, particularly in the context of central European dry grass- or heathland. Herding is a common practice in conservation grazing (Bailey et al.
2019), which avoids fencing and therefore also a direct exclusion of larger wildlife. However, shepherds and their dogs may be perceived as predators and possibly trigger avoidance behavior in red deer. A number of studies reports that they avoid areas with a high wolf predation risk and withdraw to more sheltered areas (Wolff and Horn
2003; Creel et al.
2005; Hernández and Laundré
2005) and a study by van Beeck Calkoen et al. (
2022) showed that risk effects of human activities can even outweigh those of predators.
On the other hand, there is evidence that red deer are able to adapt to predation risk by altering their temporal use of certain areas rather than avoiding them completely (risky times hypothesis vs. risky places hypothesis, Creel et al.
2008). They have also been found to habituate or adapt to human activities, as well as spatially and temporally evade different kinds of disturbances (Thompson and Henderson
1998; Sibbald et al.
2011; Westekemper et al.
2018). For example, a study by Edge and Marcum (
1985) observed that they avoid areas during ongoing logging operations and return on weekends when logging is paused. Similarly, they have been found to avoid areas close to hiking trails during daytime, when they are more frequented while returning at night (Marion et al.
2021). This ability to adapt to disturbances seems to primarily occur whenever these follow a regular pattern, but it is not evident when they are not predictable (Knight
1980; Westekemper et al.
2018). Despite this ability to adapt to disturbances, several studies acknowledge red deer site-fidelity (Switzer
1993) or spatial memory (Fagan et al.
2013) as important factors in habitat selection, meaning that they show a tendency to use territories with which they are familiar (Wolf et al.
2009; Gautestad et al.
2013). Conversely, this could mean that once they have adapted their spatial behavior to a predictable and long-lasting disturbance, they will continue to display this altered behavior in excess of the actual disturbance (e.g., Sibbald et al.
2011). Additionally, persistent scents, especially those of dogs, could negatively affect the attractiveness of these areas for extended periods (Chabot et al.
1996; Elmeros et al.
2011). To date, the scientific literature has insufficiently covered conservation grazing and its direct and indirect influences on mammalian wildlife. Particularly, very little information is available regarding herded sheep or goats and wild red deer (but see Hester et al.
1999).
In this study, we investigate the effects of conservation grazing with herded sheep and goats on the spatio-temporal behavior of wild red deer in dry heathland. In particular, we are interested in displacement effects, the temporal scale of such potentially time-lagged effects, and any signs of either adaption or habituation to conservation grazing. Specifically, we hypothesize that (i) red deer are expected to avoid areas used by livestock, as well as adjacent areas, when livestock are present, albeit (ii) red deer increase the use of these areas when sheep and goats are temporarily absent and (iii) there is a time-lagged disruption in red deer spatial behavior when conservation grazing practice ends. A potential displacement of red deer by targeted sheep-grazing—either due to resource competition or direct disturbance—would be highly relevant for both, wildlife management and conservation of semi-natural open areas.
Discussion
Our modelling approach highlights the significant effects of conservation grazing on the spatio-temporal behavior of red deer, leading to their temporal displacement from conservation grazing sites and adjacent areas up to a distance of 3000 m. Following the start of conservation grazing, the use of these sites by red deer decreases not only during the day, but also during the night and at twilight when sheep are temporarily absent; it also remains low during the first three weeks after sheep-grazing ceases. Our seasonal resource selection functions indicate that red deer regularly use conservation grazing sites and their surroundings at Glücksburger Heide during the day and at twilight in both summer and winter, but this use is reduced considerably during conservation grazing and the three weeks following these grazing activities.
This suggests that sheep-grazing with herded sheep is perceived by red deer as a direct disturbance. A wandering flock of sheep, including a shepherd and sheepdogs, represents a complex combination of multiple visual, acoustic, and olfactory stimuli. Sheepdogs especially move quickly and change direction often when working with sheep, and so their movements are difficult to predict and they are most likely perceived as a predatory threat by red deer. But also walking humans have been reported to represent a significant source of disturbance (Stankowich
2008). Moreover, the open areas used for sheep-grazing provide little cover for red deer, and it is known that they are specifically sensitive to disturbances in open areas (Stankowich
2008; Jayakody et al.
2011). Consequently, they may avoid these areas during conservation grazing, due to a combination of this disturbance and a lack cover in which to hide. Our finding that the effects of conservation grazing reach beyond the limits of the actual grazing sites into more covered areas suggests that acoustic and probably olfactory stimuli also affect red deer regardless of the available cover. An alternative explanation for these far-ranging effects could be that red deer select other open areas as a consequence of displacement, thereby resulting in a general spatial shift of their home ranges (Peek et al.
1982; van Dyke and Klein
1996). Edge and Marcum (
1985) report similar far-reaching displacement effects of logging operations on red deer, in that they remained a mean distance of 2000m away from logging operations and did not move closer than 500 to 1000 m. One additional indirect effect of conservation grazing on red deer could be related to wolves present in the study area. These might be attracted by the livestock leading to an increased presence of wolves in the vicinity of conservation grazing activities, which in turn increases general vigilance of red deer and triggers avoidance of these areas (van Beeck Calkoen et al.
2021).
The effects observed by us appear to be more pronounced in winter than in summer. The initial use of the sites indicated by logistic regression, as well as the selection of these sites and their surroundings implied by the RSF, were greater during winter. We assume that the large open heathland areas at Glücksburger Heide play a more critical role as grazing sites for red deer in winter, as there are fewer alternative agricultural food sources around the study area at that time of the year. The increased hunting pressure, especially in the surrounding areas, could also be a contributing factor. The importance of heathland areas as foraging grounds for red deer in winter is also underlined by the findings of Riesch et al. (
2019), who observed significantly greater forage removal rates in heathland at this time of year. Thus, the overall impact of sheep-grazing on red deer in Glücksburger Heide can be assumed to be greater in winter. Nevertheless, we must be careful when interpreting the statistical output of our winter logistic regression. The patterns seem less clear with generally larger p-values (Table
3) as significances in this model might be restrained by the smaller sample size of the winter models (Demidenko
2007).
Although other studies mention red deer habituating to disturbances (Thompson and Henderson
1998; Found and St. Clair
2016), we found no indication for such habituation during periods of conservation grazing with sheep and associated disturbance stimuli. However, it should be noted that we only tested for a short-term habituation effect during conservation grazing in winter. During summer, where we also analyzed the spatio-temporal behavior of red deer during later stages of conservation grazing, displacement effects increased or remained stable rather than decreased – as one would assume in the case of habituation. Both the direct use of the sites and the selection of their surroundings (RSFs) decreased at all times of the day (day, night, twilight) when conservation grazing started and continued (summer). These effects seems to be relatively less pronounced during the night suggesting direct disturbance effects of conservation grazing during daytime; but this could also indicate a general shift to increased nocturnal activities due to generally higher disturbance levels (Ensing et al.
2014). We found no indication of any short-term spatio-temporal adaption to disturbance as described by Edge and Marcum (
1985) in the case of logging activities, or by Westekemper et al. (
2018) and Marion et al. (
2021) in the case of human recreational activities. There were also no signs of any compensatory use of the conservation grazing sites by red deer at night or at twilight, when sheep were not present. Our findings are in line with those of Sibbald et al. (
2011), who did not observe any compensatory use of areas around hiking tracks during the nightly absence of hill-walkers. There are several possible explanations for this observation in the case of conservation grazing at Glücksburger Heide. We suspect that, in addition to any direct disturbance effects, sheep-grazing temporarily reduces the attractiveness of affected areas for red deer. Stewart et al. (
2002) report correspondingly that red deer avoid areas for at least seven days where cattle have grazed. Forage depletion caused by feeding and trampling sheep might be one reason for reduced attractiveness in this case (Bakker et al.
1983; Jauregui et al.
2009), but another cause might be lingering olfactory stimuli, since Chabot et al. (
1996) found that the scent of sheep and humans can reduce the general palatability of vegetation for red deer, while canid (
Canidae) scent can even provoke symptoms of physical stress or lead to avoidance of the affected areas (van Beeck Calkoen et al.
2021). This might be a crucial factor in why red deer also reduce their use of these sites during the night and at twilight when sheep are not present. An additional explanation might be spatial memory, which is an emerging field in behavioral ecology (Fagan et al.
2013). In this regard, there is evidence that spatial memory plays an important role for the use of resources by red deer (Gautestad et al.
2013), and it is likely that it also plays an important role in the avoidance of disturbances. It might even outweigh a red deer’s ability to assess the temporal dynamics of such, especially if they are temporally difficult to predict. As a consequence, they will avoid the affected areas for a certain time, even if the reason for avoiding it in the first place is no longer present.
Within the considered time period (21 days), red deer do not resort to their original spatio-temporal behavior after conservation grazing stops. Their use of the conservation grazing sites remains low at all times of the day in summer but partially recovers during the daytime in winter. We make a similar observation for habitat selection in terms of distance to conservation grazing sites: Selection is low after conservation grazing ceases during the summer but partly recovers in the winter. One important factor could be the overall greater selection of open areas by red deer, leading to faster re-utilization in winter. However, the observed seasonal differences might also be induced by additional factors such as the removal of tree saplings during summer or increased hunting pressure during winter, as well as varying amounts of food resources in the surrounding areas throughout the year. We conclude that the effects of conservation grazing on the spatio-temporal behavior of red deer in Glücksburger Heide persist for some time after conservation grazing activities stop. However, the design of this study does not allow us to make clear inferences about the exact temporal dimensions of enduring behavioral changes. Nonetheless, it should be noted that during the following 21 days, both the direct presence on the sites and the overall habitat selection significantly differ from what is observed before conservation grazing. This delay in revisiting these areas might be caused by the already mentioned depletion of forage vegetation, remaining scents, or persistent spatial memory of the disturbance, the latter of which might push these animals to other suitable areas they continue to use after sheep-grazing has stopped. Aspects of spatial familiarity (Piper
2011) might also play a role, as red deer have been found to prefer areas with which they are familiar and return to well-known foraging grounds on a regular basis (Wolf et al.
2009).
Our findings are supported by good results during the two different cross-validation approaches. The resemblance of the overall patterns between the different seasonal models provides additional confidence that our results reflect the veridical effects of conservation grazing on the spatio-temporal behavior of red deer at Glücksburger Heide. Nevertheless, the structural composition of a landscape and the overall availability of different land-cover types can essentially influence general spatial behavior (“functional responses”): Godvik et al.
2009; Matthiopoulos et al.
2011 and thus also responses to disturbance (Hebblewhite and Merrill
2008). We, therefore, suspect that the impact of conservation grazing on red deer is variable and depends on the surrounding conditions.
Our findings indicate that interactions between wild and domesticated herbivores should be considered in conservation planning for semi-natural open areas such as dry heath- or grasslands. Both conservation grazing with domesticated herbivores (Dostálek and Frantík
2008; Jauregui et al.
2009) and grazing by red deer (Riesch et al.
2019,
2020) have been found to increase plant species diversity, in order to reduce tree encroachment and to maintain the open character of such areas. In this study, however, we demonstrate that conservation grazing with sheep herding– at least temporarily – is capable of displacing wild red deer. As a result, the beneficial effects of wild ungulate herbivores and conservation grazing might compete when applied simultaneously. On the other hand, mixed grazing regimes have on other occasions proven to be especially efficient and to promote biodiversity (Rosa García et al.
2013; Fraser et al.
2014; Marchetto et al.,
2021). This could also be the case for a combination of sheep and wild ungulate herbivores such as red deer. In the anthropogenic landscape, where red deer can be attracted by other food sources such as surrounding agricultural areas, grazing with shepherded sheep can be applied in a more targeted manner. In contrast, wild red deer are less restricted by property and management boundaries and can help to counteract encroachment at forest edges and other transitional landscape elements. Other relevant consequences of the displacement of red deer from the conservation grazing sites might be related to aspects of wildlife management. It is very difficult to precisely predict any population-level effects for red deer. Agricultural areas in the surroundings probably provide sufficient though temporally shifting food sources. However, the (temporary) loss of relatively undisturbed open foraging grounds might increase red deers’ perception of risks and flight responses - especially for females or groups with young offspring (Stankowich
2008). Since the displacement effects seem to reach beyond the conservation grazing sites and affect larger areas we also expect spatial shifts in human-wildlife conflicts with forestry (Bobrowski et al.
2020), agriculture (Walter et al.
2010) or traffic (Mysterud
2004) and hunting activities might have to be adapted. All these aspects need to be considered and pondered in order to optimize the outcome. One solution could be the careful timing of grazing with herded livestock to reduce displacement of wild ungulate herbivores. In order to optimize mixed grazing schemes better mechanistic understanding of the effects of conservation grazing on wild herbivores is needed. Therefore future studies should aim to disentangle the effects of the presence of domesticated animals, dogs, and shepherds, as well as of resulting resource depletion.