Abstract
Voles can reach high densities with multi-annual population fluctuations of large amplitude, and they are at the base of large and rich communities of predators in temperate and arctic food webs. This places them at the heart of management conflicts where crop protection and health concerns are often raised against conservation issues. Here, a 20-year survey describes the effects of large variations of grassland vole populations on the densities and the daily theoretical food intakes (TFI) of vole predators based on road-side counts. Our results show how the predator community responds to prey variations of large amplitude and how it reorganized with the increase of a dominant predator, here the red fox, which likely impacted negatively hare, European wildcat and domestic cat populations. They also indicate which subset of the predator species can be expected to have a key-role in vole population control in the critical phase of low density of grassland voles. Our study provides empirical support for more timely and better focused actions in wildlife management and vole population control, and shows why it implies evidence-based and constructive dialogue about management targets and options between all stakeholders of such socio-ecosystems.
Introduction
The relationship between people and rodents is an old one and early accounts show clearly that rodents were a destructive agent in the fields and a source of disease for many ancient and current societies [1–4]. Voles can reach high densities with multi-annual population fluctuations of large amplitude, and where they occur they are mainly considered as pests. However often persecuted for this reason [4, 5], their effects on biodiversity is of crucial importance. They are at the base of temperate and arctic food webs, maintaining large and rich communities of predators, and modifying nutrient cycling, soil aeration, and micro-organism assemblages [6]. This places them at the heart of management conflicts where crop protection and health concerns are often raised against conservation issues [5]. Moreover pest control, especially when using chemicals carelessly, by indirect poisoning, can non-intentionally depress the populations of predators able to contribute to the regulation of rodent populations [7, 8].
Predation has been suggested as one of the main drivers of rodent population fluctuations. Theory predict that specialist predators that feed on one or a few kinds of prey can destabilize prey populations because they exert delayed- and direct density dependent mortality on their prey populations, while generalist predators, that feed on a wide variety of prey species, have direct density dependent mortality and therefore stabilize prey populations [9]. However, experimental tests of this prediction, e.g. predator removals and comparative field studies, have provided evidence both supporting and rejecting this hypothesis [10–15]. One of the world’s simplest vertebrate predator/prey community is high-arctic tundra ecosystem. Its has only four predators preying upon one rodent species, the collared lemming (Dicrostonyx groenlandicus). There the numerical response of the stoat (Mustela erminea) drives the population dynamics of the collared lemming by a 1-year delay. this dynamics is concurrently stabilized by strongly density-dependent predation by the arctic fox (Vulpes lagopus), the snowy owl (Bubo scandiacus), and the long-tailed skua (Stercorarius longicaudus) [14, 16]. Studies in Fennoscandia on small mammal population cycles have accumulated support for the predation hypothesis [17]. Population dynamic patterns of the common vole (Microtus arvalis) in intensive agricultural landscapes of south-west France are largely consistent with five of six patterns that characterize rodent cycles in Fennoscandia and can be explained by the predation hypothesis [18]. Hence, there is little doubts that in European arctic and temperate ecosystems predation plays a key role in regulating small mammal population dynamics. However, in temperate ecosystems the multiplicity of prey-resources and the larger number of predator species combined to landscape diversity (e.g. the spatial arrangements of optimal and suboptimal habitats for prey and predators) [19, 20] make the disentangling of the detailed processes and the role of each species involved still a challenge [21].
Interactions between populations of various species in systems driven by predation are complex. For instance, based on a 20-year survey of the effects of an epidemic of sarcoptic mange decreasing fox populations in Scandinavia, Lindström et al. [22] revealed that red fox (Vulpes vulpes) predation was a crucial factor in limiting the populations of hare (Lepus europeus), capercaillie (Tetrao urogallus), and grouse (Tetrao tetrix and Bonasia bonasia), as well as fawns per doe of roe deer (Capreolus capreolus) in autumn, and in conveying the 3-4 year fluctuations of voles (both bank and field voles (Myodes glareolus and Microtus agrestis)) to small game. The importance of such prey switchings on prey population dynamics has also been reported since long in northern ecosystems of the Newfoundland, in Canada, where lynxes (Lynx lynx), prey on snowshoe hares (Lepus americanus), until the hare population crashes. Then, lynxes switch to caribou calves (Rangifer tarandus), and the cycle continues [23]. Furthermore, variations in prey and predator population densities also affect parasite transmission and shape disease epidemiology as shown e.g. for the common vole and tularemia in Spain [24] or the montane water vole (Arvicola terrestris), the red fox and Echinococcus multilocularis in the Jura massif, France [25, 26]. As a whole, those multiple and complex interactions can hardly be investigated in depth by simple modelling [27] or by small-scale experiments that cannot technically take into account all the relevant space-time scales and species involved in the real world and thus be generalized.
However, stakeholders in such systems are often protagonists of endless debates about regulation adoption and management decisions, each of them advocating for the control of one among many possible of the population targets and subsequent options for management. This debate is recurrently illustrated by controversies about large and mesopredator culling (e.g. wolf, fox, etc.) for ‘desirable’ prey protection (e.g. sheep, hare, etc.) as opposed to their protection for their role as regulator of ‘undesirable’ prey such as overpopulating cervids or small mammal pests [5, 28]. Due to the fascination they exert over the general public, the control of large or medium carnivores is generally unpopular whatever the reasons. By contrast the broad social acceptability of the control or exploitation of other species (e.g. rodents, lagomorphs), due to their lack of charisma or low public visibility, and the lobbying of pro/anti, may lead to social consensus or political decisions that hinder evidence-based and targetted management options, and can even jeopardize biodiversity conservation (e.g. the protection of species of conservation value critically endangered sometimes threaten by prey switches) [29].
Instruments to help set local, national and global priorities in biodiversity conservation, pest and game management and disease control can be severely handicapped by a lack of sound observational data, collected through fieldwork [30], especially in complex multi-species and multi-functional systems where modelling fails (e.g. theoretical sub-systems oversimplified for mathematical convenience or lack of relevant external validation) and where experiments cannot be performed on space-time scales relevant for many of the species involved (large home-range, slow growth rate, etc.) for obvious practical and sometimes ethical reasons. However, our knowledge of the ecology and population dynamics of many common species and communities is still limited. Field-based investigations, including somehow despised ‘observational’ or ‘correlative’ studies are powerful allies of synthetic and comparative studies [30], providing crucial data that enable us to better identify key-factors driving those systems and inform decision-making across scales, design tests and modelling at the appropriate place and time, and make adaptive management, necessarily linked to adaptive monitoring [31], possible.
Such an approach has been successfully carried out in the Jura massif for small mammal pest control [32]. In this region of mid-altitude mountains (250-1718m), in the 1950s, grassland was covering only 20-60% of farmland whatever the altitude. Farmers in higher altitude areas (above 500 m a.s.l) specialized in milk production in the 1960s for the production of a number of Protected Geographical Indication cheese (Comté, Morbier, Mont d’Or, Bleu de Gex). This led to convert ploughed fields into permanent grassland and subsequently to increase the proportion of permanent grassland up to 75-100% of farmland in these areas, with, additionally, an increase in grass productivity from 2-3 tonnes of dry matter.ha−1.year−1 in the 1950s to 4-9 tonnes nowadays (milk production itself shifted from less than 2000 liters.cow−1.year−1 in 1951 to currently about 7000 liters in average (range 5400-8500)). As soon as the 1970s, this shift resulted in triggering massive outbreaks of a grassland vole species, the montane water vole, with 5-6 year cycles propagating over the grasslands of the Jura massif under the form of a travelling wave [33, 34]. In the same area, it also favoured outbreaks of the common vole, another grassland vole, however non-cyclic [6]. A number of field studies and modelling have shown that population dynamics of the two species was shaped by landscape features, with hedgerow networks and wood patches dampening the population dynamics and by contrast open grassland landscapes amplifying the outbreaks [33, 35–40]. Those observations permitted to shift rodent control from late and chemical-only (with devastating side-effects on non-targeted wildlife) in the 1980-1990s [41], to early and more environmentally-friendly multi-factorial control in the 2000s [7, 8] using an integrated approach based on the key factors previously identified in observational studies [32]. The successful results of this shift in practices with regard to pest control can be considered as a quasi-experiment, and confirmed that the correlations identified in earlier studies were not spurious. Those field studies also indicated that the population dynamics of all of the rodent species in the area were synchronic, with concomitant low density phases [42, 43]. Hence, after a tipping point in the 1960s from mixed agriculture to specialized milk production, the regional socio-ecosystem stabilized since the 1970s, with cyclic outbreaks of montane water vole and outbreaks of common vole. Those outbreaks provide regularly massive quantity (up to > 8Okg.ha−1) of prey for carnivores and birds of prey in grassland and by contrast low densities of secondary prey-resources less accessible (vegetation and/or anti-predation behaviour) like forest, marsh and fallow small mammals (maximum about 3kg.ha−1) (e.g. bank vole, wood mice, (Apodemus sp.), field vole, etc.), with periodic (5-6 years) concomitant low densities in every habitats.
If landscape effects on vole populations have been assumed by deduction to be in part mediated by predation [20, 35, 44, 45] in this unique system, some field and modelling studies have focused on the functional response of predators, e.g. the barn owl (Tyto alba) and the red fox [27, 46–48], but none on their population dynamics. However, predator management generates controversy regularly in such system, where their role and the ecosystem services/disservices [49] they render are perceived like contrasted. Socially assumed to contribute to small mammal pest regulation, they are also blamed for maintaining other species in predation sinks, such as small game or some species of conservation value vulnerable to predation (e.g. wading birds, capercaillie, etc.), for damage to hen house, etc. However, observations to corroborate those perceptions are often anecdotal, often with no context, and call for sound data on which objectively ground adaptive management. Although the density of small and medium-size mustelids can hardly be estimated using simple techniques, other predator species can more easily be censused by road-side counts. The variation in this predator community structure over the time span of large fluctuations of prey abundance has not been documented yet in this system, limiting both comparisons with ecosystems described in other part of the world where small mammal outbreaks occur [4] or with more simple food webs of northern ecosystems. Moreover here, a large scale inadvertent experiment was offered by the chemical control of vole populations in the 1990s, leading to a dramatic decrease of the fox population and its gradual recovery the following years after a shift in vole control practices [7].
The aim of this 20-year study is to describe the effects of large variations of grassland vole populations on their predator community and of the long term increase of the fox population in such system. It aims at understanding (i) how a predator community respond to prey variations of large amplitude, (ii) how this community reorganizes with the increase of a dominant predator, here the red fox, (iii) which subset of the predator species can be expected to have a key-role in vole population control in the critical phase of low density of grassland voles. Better understanding the links between grassland vole population variations and predator responses will allow more timely and better focused management actions for all stakeholders in multifunctional socio-ecosystems.
Material and methods
Study area
The study was carried out around the Pissenavache hamlet (46.95°N, 6.29°E) in Franche-Comté, France, in an area of 3425 ha (2646 ha of farmland, 1094 ha of forest, 167 ha of buildings), at an average altitude of 850-900 m above sea level (Fig. 1 and 2). There, 100% of the farmland was permanent grassland used for pasture and (high grass) meadow for cattle feeding in winter (minimum 6 months, November-March), with a productivity ranging 5-6.5 tonnes of dry matter.ha−1.an−1 under the specifications of the European Protected Geographical Indication of the locally produced Comté cheese. A KML file with the bounding box of the study area is provided in Supplementary Material S1 kml file.
Road-side counts
Predator and hare (Lepus europeus) populations have been monitored from June 1999 to September 2018 (20 years) using night and day road-side counts. Each sampling consisted in driving a car with 4 people (the driver, a data recorder and two observers) along a fixed track at less than 20 km/h. The length of the track was 18.6 km from 1999 to 2009, then 19.6 km due to a slight variation in the itinerary (Fig. 1). Observations were made using 100-W spotlights by night and binoculars for species identification. Distinction between domestic cats (Felis silvestris catus) and European wildcats (Felis silvestris silvestris) was made visually considering phenotypic criteria (relative to pelage and morphology) without possible distinction of hybrid individuals. Double counting was unlikely because transects were relatively straight, space open (Fig. 2) and observers were careful about animal movements. Sampling was carried out on 3 successive nights (4 when meteorological conditions prevented to achieve a sampling) called a ‘session’. The same track was also driven by daylight in the early morning.
Most often 3-4 sessions a year were carried out corresponding to seasons, but 1 session in autumn only since 2016. Day road-side counts were stopped in 2017 (see Supplementary Material S2 Excel file for details). Each observation was recorded on a paper map (IGN 1/25000). A Kilometric Abundance Index (KAI) was calculated for each session as the maximum number of animals recorded km−1 (thus providing a lower limit for the number of animals present). For the period 2001-2006, only the total counts without the localizations of the observations were available. Thus, only the 1999-2000 and 2007-2018 observations could be georeferenced in 2019.
Small mammal relative densities
Transects
Small mammal (A. terrestris, M. arvalis and Talpa europea) relative abundance was assessed using a transect method adapted from [50–52]; a 5 m-wide transect across the study area was divided into intervals 10 m long and the proportion of intervals positive for fresh indices (tumuli, molehill, runway, feces, cut grass in holes) was taken as an index of abundance. Total transect length was 11.6 km (Fig. 1). Sampling was carried out once a year in april 2007, then in august from 2008 to 2010, then at least twice a year generally in spring and in autumn from 2011 to 2018 (see Supplementary Material S3 Excel file for details).
A. terrestris communal scores
In order to get abundance assessments on a larger space-time scale, abundance was also assessed at the commune-scale by technicians of the FREDON of Bourgogne Franche-Comté (a technical organization for plant pest prevention and control contracted by the Ministry of Agriculture [53]), in the 7 communes crossed by the road-side count itinerary (Fig. 1). Assessments were made in autumn since 1989 (Supplementary Material S3 Excel file). The FREDON assessment uses a ranking system that ranges from 0 to 5: 0 - no A. terrestris sign in any parcel within the commune; 1 - low or no A. terrestris tumuli, voles and moles (T. europea) cohabiting the same tunnel systems; 2 - A. terrestris tumuli present in some parcels within the commune and mole burrow systems still present in some parcels; 3 - A. terrestris tumuli present in some parcels within the commune, few or no mole burrow systems present in the commune; 4 - A. terrestris colonies established in the majority of meadows and within pastures; 5 - all of the commune is colonized by A. terrestris. The FREDON index does not directly translate to transect-based indices, partly because it is applied at the commune scale and not the parcel scale, but Giraudoux et al. [52] have shown that levels 0-1 correspond to densities < 100 voles.ha−1, level 2 to 100-200 voles.ha−1, and levels 3-5 to > 200 voles.ha−1. For a given year, the median score of the 7 communes was taken as a score of abundance.
Grassland prey resource
In order to better visualize grassland rodent populations variations on the same scale and fill the gap when transects data were lacking, taking into account that M. arvalis body mass is four times smaller than A. terrestris’s in average [54], the dynamics of prey resource abundance in grassland has been indexed (i) over the time span when transects were carried out, summing the relative abundance of A. terrestris and of M. arvalis divided by four, divided by the maximum of this sum over the series and (ii) before this time span, using the FREDON score divided by its maximum score (5). The amplitude of the high density phase is biased to an unknown extend with this method (e.g. arbitrarily summing weighted relative abundances, chained with standardized FREDON scores), but not the time-locations of the low density phases. Thus, the alternation between high density and low density phases was robustly and correctly represented over the time series as an abundance index, in the best possible way given the data, for further comparisons.
Complementary data
In France, bromadiolone, an anticoagulant rodenticide, has been used to control water vole populations since the 1980s, with deleterious effects on non-target wildlife including vole predators [8]. In the early 2000s, the development of an integrative pest management (IPM) approach [32] led to decrease dramatically the quantity of bromadiolone applied by farmers and their non-intentional effects [7, 8]. By law, the delivery of bromadiolone baits for vole control to farmers is under strict FREDON supervision and usage declaration compulsory in order to ensure traceability [55]. Data on bromadiolone quantities used in the 7 communes of the study area were provided by the FREDON of Bourgogne Franche-Comté.
Theoretical daily food intakes (TFI) per predator species were computed following Crocker et al.’s method [56] with small mammals considered as prey. Average body mass when missing was estimated based on the Encyclopédie des carnivores de France [57–60], the Handbook of Birds of Europe, the Middle East and North Africa [61] and the Encyclopedia of Life (https://eol.org).
Statistical analyses
Statistical and spatial analyses were performed in R (version 3.6.2) [62] with the packages Distance [63], pgirmess [64], rgdal [65], rgeos [66], using QGIS 3.10 [67] complementarily. The standard errors of small mammal relative abundances assessed from transects were computed across 1000 bootstrap replicates [68]. We examined the effect of grassland prey abundance indices and seasons upon the number of animals observed in the road-side counts using generalized linear models with a Poisson error distribution of the form n = a0 + a1ln(x1) + a2x2 + a3x3 + ϵ, with n, the number of observations, x1, the length of the itinerary, x2, the season, x3, the prey abundance index, ai, the model coefficients, ϵ, the residuals. To avoid the overestimation of degrees of freedom that might come from time series data (here irregular and intrinsically autocorrelated), statistical inference was computed using permutation tests. The grassland prey resource index corresponding to each road-side count was linearly interpolated over time between the two bracketing abundance index estimates.
The shortest distance of observations to the road-side count itinerary, to the nearest forest and to the nearest building were computed [65, 66] and their distribution examined. In order to test whether the proximity of some habitats might explain the observed distributions, mean distance to forest and to building were compared to mean distances obtained from 1000 simulations of the same number of random positions as the number of observations in the strip observed along the itinerary.
In order to obtain density estimates, distance to the itinerary data were analysed using conventional distance sampling with a truncation distance [69–71] including 90% of the observations for each species at the minimum. Avoidance behaviour along the road being detected for most species, we used hazard-rate detection functions fitted to the data. This function type has a more pronounced shoulder compensating for the bias due to avoidance [63]. Models with a seasonal effect as covariate were compared with concurrent models with no covariate using the Akaike Index Criterion [72].
Results
Small mammal density and prey resource variations
Fig. 3a shows the cyclic variations of A. terrestris from 1989 to 2018. Predators communities have been monitored during the last four cycles, but the local populations dynamics of small mammals during the last three cycles only (Fig. 3b). A clear synchrony of the low density phase between rodent species was observed, while T. europea and A. terrestris peaks were in phase opposition. In term of prey resource, low density phases contrast with the phases of large abundance of grassland voles (Fig. 3c).
Time variations of predator and hare relative abundances
Twenty seven species for the day road side counts and 24 for the night were observed, corresponding to 19,010 and 7,355 individual observations respectively, and to 58 sessions for each count type (≃ 348 night or day counts in total) (see Supplementary material S2 Excel file). Some were occasional visitors and likely play a marginal role on vole prey (e.g. grey herons (Ardea cinerea) could regularly be observed preying on voles in grassland). Others, such as some mustelids (stoat (Mustela erminea), least weasel (M. nivalis), stone marten (Martes foina), pine marten (M. martes)) were elusive and hardly detected by road-side counts. Among them, the following species were both observed frequently enough over time and considered of interest for this study: for day road-side counts, the carrion crow (Corvus corone), the common buzzard (Buteo buteo), the red kite (Milvus milvus), the kestrel (Falco tinnunculus), the domestic cat (Felis silvestris catus), the hen harrier (Circus cyaneus); for night road-side counts, the European hare (Lepus europeus), the red fox (Vulpes vulpes), the domestic cat (Felis silvestris catus), the European wildcat (Felis silvestris silvestris), the long-eared owl (Asio otus), the European badger (Meles meles).
Fig. 4 shows the dynamics of diurnal species. For each species KAI differences between seasons were found statistically significant except the domestic cat (Table 1 and Fig. 5).
For instance, common buzzard KAI was highly significantly correlated to grassland prey index, with KAI 2.2 times higher in autumn than in spring. In spring, during the breeding season, KAI were 4.3 times larger in the peak phase than in the low density phase of grassland vole populations. Red kite’s correlation p-value was equal to and kestrel and hen harrier’s above but not far from the critical threshold generally accepted of p(Ho) ≤ 0.05. This lack of significance for the latter two species held from one outlier, when a relatively large number of birds was observed in Summer 1999 in a period when prey estimates were not available locally from transects but derived from the FREDON scores on a communal scale. Dropping this observation from the data set would lead to reject Ho at p = 0.01 and p = 0.02, respectively, and to conclude formally on a correlation between the number of observations of those species and grassland prey abundance.
Fig. 6 shows the dynamics of nocturnal species. We did not detect numerical response statistically significant to seasons or grassland prey index for red fox, badger and long-eared owl abundance. Domestic cat did not correlate to grassland prey index but to seasons, with lower counts in winter. Hare and wildcat KAIs were significantly correlated to grassland prey index but seasonal variations could not be detected (Table 1 and Fig. 7). Fox and hare KAIs were highly and negatively correlated to each other (p < 0.001). Furthermore, a model of hare abundance as response variable including grassland prey index and fox KAI as independent variables showed that controlling for grassland prey, hare abundance did not significantly correlate to fox KAI at a probability ≤ 0.05 (however with an observed p-value of 0.07).
Red fox and badger showed significantly higher abundance in average in the last half of the time series, and hare, wild and domestic cat, long-eared owl and hen harrier significantly lower (one-tailed permutation tests on mean, p < 0.001) (Fig. 4 and 6).
Spatial variations
Observations were truncated at a distance of 300 m and 350 m from the track for night and day road-side counts respectively, accounting for 92% and 93% of their total number. Among all species in the open grassland strip along the itinerary, only the common buzzard with regard to forest and buildings, and the red fox with regard to buildings were randomly distributed. Carrion crow, red kite, kestrel and hare were observed at a greater distance to forest than expected from a random distribution; hen harrier, red fox, wildcat, long-eared owl, badger at a smaller distance; wildcat, long-eared owl and badger at a greater distance to buildings; carrion crow, red kite, kestrel, domestic cat, hen harrier at a smaller distance (Tab. 2). 75% of the observations of domestic cat were made at less than 500 m of buildings by night and at less than 250 m by day (Fig. 8). No change in any of those patterns was observed between the first and the second half of the time series.
Estimation of predator and hare densities, and theoretical food intake (TFI)
Comparing detection models with ‘season’ as covariate with models with no covariates led to reject the hypothesis of a seasonal effect on the detection function for every species (detection functions are presented in Supplementary material S4 Word file and S5 Word file). Based on the visual examination of KAI dynamics, for each species, we identified periods when the indices could be considered similarly high or similarly low with regard to the amplitude of variations and categorize them as sub-samples of ‘low’ or ‘high’ densities (see Fig. 4 and 6). Table 3 gives conversion coefficients from KAI to densities, the maximum density values observed, and summarizes the estimations obtained by distance sampling by density categories (’low’ or ‘high’). Considering the relative aggregation of the domestic cat close to buildings, we provide one density estimate for the entire study area, and another for a buffer of 300 m (night) or 250 m (day) around buildings.
Fig. 9 shows the population density variations of the predator community during the study period for all species when both day and night road-side counts were available. Raw numbers and graphs on biomass and daily TFI variations are provided in Supplementary material S6 Excel file, S7 Word file and S8 Word file. The main features of the dynamics hold in the importance of the carrion crow (range 4.4-56.9% of the total TFI), the common buzzard (range 4.7-48.6% of the total TFI) and the red kite (0-54.5% of the total TFI) along the entire time span, and the gradual increase of the red fox from 1999 to 2010 (Fig. 9). The numerical importance of the carrion crow apart, three key periods could be identified: 1999-2004 with red fox density extremely low not exceeding 0.2 ind.km−2, the community being numerically dominated by cats (domestic and wild) and common buzzards, 2005-2009 with an increasing density of foxes, and 2010-2016 when fox densities were higher and stabilized at an average of 2.7 ind.km−2. Foxes represented only 5.5% of the predator biomass (2.8% of the total TFI) in 1999-2004 but reached 29.5% (31.4% of the TFI) in 2010-2016. Whatever the period and relative densities of species, the average TFI in the three periods was close to 4 (3.8-4.2) kg.km−2.day−1. The largest predator densities were reached during the high density peaks of the grassland vole populations, with a maximum observed in the autumn 2008, with 60 ind.km−2 (carrion crow making 48% of this total) and a daily TFI of 10.7 kg.km−2.day−1 (39.3% from carrion crow).
Table 4 summarizes results at the grassland vole populations peaks of the autumns 2003, 2008 and 2012, and at the low density phases of autumn 1999, spring 2007, autumn 2010 and spring 2014. In the autumn 1999, the first 4 species totalling 91% of the TFI were the carrion crow, the common buzzard, the domestic cat (night) and the kestrel. The common buzzard was still among those first four species in the next low density phase (spring 2007), but the proportion of TFI from birds of prey still decreased, and was preceded by the fox, the carrion crow, the domestic cat and the wildcat in autumn 2010 and spring 2014, these species making together 86% and 84% of the TFI. However, a large proportion of domestic cats roaming at less than 500 m of the buildings, far from villages where domestic cats were virtually absent, fox, carrion crow and wildcat alone made 86% of the TFI. During the first two high density phases, the carrion crow, the common buzzard, the red kite and the domestic cat (night) made 81 and 91% of the TFI, and in autumn 2012 during the third high density phase, the fox, the common buzzard, the carrion crow and the domestic cat (day) alone made 81% of the TFI. Table 4 also shows that the TFI ranged between 1.5 and 2.7 kg.km−2.day−1 in the low density phases and between 6.9 and 10.7 kg.km−2.day−1 in the high density peaks, thus the TFI was multiplied by 7.1 at the maximum whilst grassland small mammal population biomass was multiplied by thousands.
Discussion
Response to grassland vole population variations
The general pattern of small mammal population variations observed in our study area confirms earlier observations over the shorter period 1987-1990 [42] for rodents, and the fact well established that A. terrestris outcompete T. europea for space [52, 73]. T. europea at its population peak does not exceed some individuals per hectare, and by distaste and quantity is a negligible prey for most predators [54].
Among the 11 predator species monitored 4 (maybe 6) show a numerical response to the large variations of grassland prey observed over the 20 years of our study. Namely, the common buzzard, the red kite, the wildcat and also the hare, and possibly the kestrel and the hen harrier. However, such response were modulated by population trends on a larger scale. This was the case for the hen harrier and the long-eared owl, with populations decreasing over time in the study area reflecting the general decrease of those species in Franche-Comté and nearby Switzerland [74]. Those variations were also seasonal with generally larger populations in autumn, but in summer for the red kite, corresponding to dispersing birds after reproduction and post-breeding migration. The numerical response of the hare, an herbivore, to grassland vole density variations is more surprising (but see next section). A similar pattern has been observed nearby at a 30 km distance from the study area, from 1976 to 1995, for the capercaillie (Tetrao urogallus), in the Massif du Risoux, where the number of fledglings per hen was positively correlated to the cyclic abundance of A. terrestris populations [75]. This response was interpreted as being the result of predation switches during the decline phase of the voles, with a predation pressure upon the capercaillie supposed to be relaxed during the high density peak, a phenomenon well documented e.g. in Scandinavian ecosystems [22, 76, 77].
The variations of the population of the other species were independent from grassland vole populations over the study time span.
Interactions within the predator community and with the hare population
A striking feature of the population dynamics observed is the increase of the fox population from the beginning of the study to the autumn 2010, independently from vole populations variations. This increase can be attributed to changes in grassland small mammal control practices by farmers who shifted from late-rodenticide-only to early-integrated control in the early 2000s [32], dividing by more than forty four the quantity of anticoagulant rodenticide used during the 2010-2018 cycles compared to 1996-2000 (Fig. 3a). Massive use of anticoagulant rodenticide is known for its deleterious side-effect on vole predators [8], with sensitivity for canids more than 3 times higher than for felids [78], and this effect has been proven to drastically decrease fox population in the area at the end of the 1990s [26]. Furthermore, Jacquot et al. [7] have shown how fox population has recovered on a regional scale after such a change in rodent control practices. In our study, the predator community shifted from a very low fox density of 0.1 ind.km−2 (CI95% 0.01-0.3) foraging in grassland up to a much larger fox abundance of 2.6 ind.km−2 (CI95% 2.2-3.2), with a peak at 4.9 ind.km−2 in the autumn 2012 (then a stabilization or a slight decrease was observed with an epidemic of sarcoptic mange, still ongoing). This value is one of the highest population densities reported in rural landscapes of Europe [79, 80]. This increase was concomitant with a sudden and dramatic decrease of the hare population during a low density phase of the vole populations, and also with a decrease in the wild and domestic cats. This strongly suggests that those declines might be the consequences of the increase of the fox population, possibly by direct predation or by creating a ‘landscape of fear’ [81, 82] limiting the distribution of the prey species to shelter-areas where they could not be detected by road-side counts (houses, forest, etc), or both. In Australia, experiments of fox removal showed in one study that cats foraged more in open habitats where foxes were removed [83] and in two others that they were more abundant [84, 85]. Furthermore, in western Poland, hare population responded the same year with 1.7 times higher density to fox removal [86], as well as positively to a sarcoptic mange epidemics depressing the fox population in Scandinavia [22]. We did not observe changes in the spatial distribution of species between the first and the second half of the study, making the ‘landscape of fear’ hypothesis less likely here, thus suggesting a major role for direct predation.
However the long-term increase of the European badger population since rabies vaccination in the early 1980s is well documented in Europe [87–89], in our study, the sudden increase since summer 2013 stays unexplained.
Except the stability of the carrion crow population at a large number, a striking feature of our system is the change in the predator community structure over the study time span. In the early 2000s, the community was numerically dominated by the common buzzard and domestic and wildcats, and with the increase of the fox population became numerically dominated by the fox itself. However, foxes did not added their number to the other predators and this population increase did not led to an increase in the average number of predators present in the study area. Large variations in vole predator number could be clearly attributed to the temporary increase of the populations of mobile birds of prey (common buzzard, red kite, etc.) in response to grassland vole outbreaks. This stability in the average predator number observed (e.g. in the low density phases of vole populations) suggests compensations among resident species due to predation or competition. Similar compensation has already been suspected in Fennoscandia, where an experimental removal of avian predators in order to understand their role on vole population regulation led to least weasel density increase [90]. In our study, the lack of data about Mustela sp. and Martes sp. does not permit to know whether those compensations observed in a community subset extend to the whole community of vole predators. Earlier studies in the area and a nearby valley of Switzerland [91, 92] and also in Fennoscandia [93] and northern Spain [15], clearly show that least weasel and stoat abundances follow grassland vole population peaks. Furthermore, small mustelid abundance has been shown to be dampened by fox in north America [94], by generalist predators [93] and by birds of prey [90] in Fennoscandia, but those interactions, possible in our study area, stay unexplored here. Moreover, small mustelids forage in vole galleries and shelter there from bigger predators. The use of rodenticide baits buried in vole galleries as enforced by the regulation [55] might contribute to an additional specific depression of small mustelid populations locally [95, 96].
Impact of predators on grassland vole prey
This study is the first one, to our knowledge, to provide data on the variations of the population densities and daily TFI of a large community of vole predators in a temperate ecosystem in response to large variations of cyclic grassland small mammals over 20 years (four A. terrestris population cycles). Several biases are inherent to the methods used (see study limitations above), however, we consider that some robust conclusions can be carefully drawn from this exceptional long-term data set. One additional limitation comes from the fact that the functional response of each species (the dietary variations as a function of available food resources) was not studied parallel to the variations of population densities. This limits the interpretation that can be given to the variations of daily TFI and the evaluation of its impact on prey populations. Thus, here we consider first what we know about predator diet, before discussing their impact on vole prey.
Dietary issues
The carrion crow is mostly opportunistic and feed principally on invertebrate, cereal grain but also small vertebrates, bird eggs, carrion, in various proportions according to place and seasons. At the extreme, vertebrate and eggs in particular can reach 86.6% of dry weight of pellets in winter e.g. in south Spain, and they are often seen to cooperate when killing small vertebrates in pair or small groups, also commonly forcing other birds including raptors to drop prey [61]. Their behaviour has not been systematically studied in our area and the importance of small mammals in the diet is not known yet, however all of the behaviours mentioned above, including scavenging on dead animals, hunting voles and forcing raptors, have been occasionally observed [97]. Thus one can hardly infer conclusions about the impact of a so opportunistic species in this ecosystem e.g. on vole regulation. Mechanically, their number however has likely a chronic impact on species vulnerable to predation such as small game and bird nests.
The other species are more specialized on small mammal prey. The detailed diet of the domestic cat is unknown in our area. However in a similarly rural area of the Ardennes, rodents make 55.9% of the dietary items found in 267 domestic cat faeces (6% birds, 36.7% human-linked food), with little difference between outdoor cats (owned by people other than farmers) and farm cats [98]. Rodents (Murids and Cricetids) constitute the main prey of wildcats, and can account for 97% of diet composition [99], while lagomorphs and birds appear generally as alternative prey. However, when the availability of lagomorphs increases, wildcats can substantially shift their diet towards them [100].
In the area, the dietary response of the red fox to variations of grassland vole relative densities differs between M. arvalis (no response) and A. terrestris (Holling’s type III-like) [48]. M. arvalis could make up to 60% of prey items in faeces even at very low density (range 0-80% of prey items on the whole range of vole densities), and A.terrestris showed a sigmoid increase with quickly a plateau (at 15% of positive intervals of a transect-see material and methods) where it made 40% of diet items in average (range 0-80% of prey items). The description of the dietary response in this context where the two main prey abundances vary among several other alternative food resources is quite complex [21, 46, 47, 101]. The comparisons of multi-species functional response (MSFR) models with empirical data about the red fox and the barn owl showed that switching between prey depends on the proportion of the prey available among other prey (frequency dependence), as commonly thought, but is also dependent on the total amount of prey (density dependence), with non-linear frequency and density dependent interactions [27].
Impact of predation on vole population dynamics
In our study area [33, 34], the population of the main prey species varied between 0 and about 1000 ind.ha−1 on a scale of tens of km2, an amplitude 5-100 times larger than those observed on a similar scale e.g. in the Greenland Arctic with D. groenlandicus (0-10 ind.ha−1) [14], in Revinge area, Sweden with M. agrestis (5-10 ind.ha−1) [93], in Messarges area, France with M. arvalis (5-120 ind.ha−1) [21], or in the Kielder forest, Scotland with M. agrestis (20-250 ind.ha−1) [102]. A similar amplitude has been reported locally for M. arvalis in alfalfa semi-permanent plots of some ha in an intensive agriculture matrix of ploughed fields of western France (50-1500 ind.ha−1) [103]. Abundance estimates of other studies were not expressed as densities, however one can reasonably assume that in crop field landscapes of Spain [104] and north-eastern and central Europe [4, 105], M. arvalis can reach also densities exceeding by far 1000 ind.ha−1 locally in grassy field margins and semi-permanent leguminous or grassland plots, but those densities are limited to a small fraction of farmland in large matrices of suboptimal habitats such as ploughed fields [44]. In our study area, two species, A. terrestris and M. arvalis had large fluctuations of similar amplitude against only one in the other systems, population surges extending on large areas of tens of km2 of highly productive and connective grassland. This ecosystem offered periodically (permanently on a large scale) an incredible biomass of several tens of kg.ha−1 of voles easy to access in grassland, to a large number of predator species. Here, we will try to understand in such system whether there are periods in vole population fluctuations when predation can be a key-factor controlling vole densities. At its maximum during the autumn 2008, the TFI was 10.7 kg.km−2.day−1, hence, with an average weight of 80 g.vole−1 [54], the equivalent of 134 montane water voles.km−2.day−1. With a carrying capacity of 1000 water voles.ha−1 and a predator diet made of an unrealistic 100% water vole at high density of voles (a food intake figure totally unlikely considering predator diets not exclusively based on voles even when specialized, see e.g. the carrion crow’s, the fox’s, etc.), above 78 voles.ha−1, this community would not be able to decrease the vole population during its growth phase (for simulations, see https://zaaj.univ-fcomte.fr/spip.php?article114&lang=en and https://github.com/pgiraudoux/shinyPred/tree/master/shinyPred_en for the code). At densities of voles exceeding some tens voles.ha−1, predators alone do not appear to be capable of instigating a population crash in our area. By contrast, daily TFI at a low or medium water vole densities can considerably slow down the population increase. For instance, with a population of 2 vole.ha−1 at the beginning of the reproduction season, and a conservative 50% of voles in the diet for the lowest TFI (1.9 kg.km−2.day−1, hence 12 voles.km−2.day−1 in spring 2007), voles would be 27 ind.ha−1 at the end of the year instead of 91 ind.ha−1 without predation. Some parameters configuration based on observed TFI at low density can even lead to vole extinction (e.g. autumn 1999 and 2010). Hence, to summarize, in our study area, the increase of predator populations due to mobile predators during the growth and high density phase of a grassland vole cycle, however responsible for consuming several ten thousands of voles.km−2.year−1 and for reducing the vole population dramatically, was likely not enough to trigger alone the decline of vole populations. However, predators during the low density phase were enough to considerably slow down the growth phase or even to extinct vole populations locally.
Furthermore, our study documents the fact that domestic cat populations could reach much higher densities of 2.4-9.1 ind.km−2 up to more than 18 ind.km−2 around villages within a 250-500 m radius, except in the winter nights when they likely prefer to stay warmly at home. In south-central Sweden, Hansson [106] observed that domestic cats, supplied with continuous alternate food, were able to dampen the population fluctuations of the field vole, compared to more or less cat-free areas. In villages at some kilometers of our study area, Delattre et al. [35, 45] reported a systematically decreasing abundance of common vole colonies around villages near our study area during similar fluctuations of vole abundance, within an area extending 300 to 400 m from the village edge, this gradient persisting throughout a complete vole population fluctuation. They subsequently hypothesized that this lower density of voles might be the result of cat predation around villages. This figure and our estimates indicate that the combination of domestic cat density and diet, added to the density and diet of the other predators, is enough to explain this effect.
The specific distribution of domestic cats, close to villages, can also cause spatial heterogeneity in predation pressure. For instance, during the small mammal low density phases, their proportion varied between 5.9% (autumn 2010) and 23.4% (spring 2007) of the total number of predators counted. In areas far from villages (e.g. > 500 m) where domestic cats were rare or absent, in reality, the real density of vole predators could be locally lower than the numbers given in Tab. 4 (e.g. 33.5 versus 43.7 ind.km−2 in spring 2007) or differences be small as in the autumn 2010 (19.7 versus 20.9 ind.km−2).
Conclusion
Overall, our results indicate that in such ecosystem with large variations of grassland prey, the structure of the predator community can change over the long term without changing its overall variations over a rodent cycle, and its TFI variation pattern. Although an unknown remains about the role of small and medium mustelid populations, the higher predator densities observed during the grassland rodent peak was mostly due to mobile birds of prey which followed the rodent population increase. However, resident predators alone during the low density phase of grassland rodent populations were capable to slow-down the increase or even to extinct rodent populations locally, but the whole predator community alone was unable to explain the population decrease observed after a high density peak. In such system, the carrion crow was numerically the largest population with the largest TFI, but its impacts in the ecosystem could not be clearly assessed due to its diet eclecticism. After a shift in rodent control practices and a much more moderate usage of anticoagulant rodenticides, the red fox population recovered and then stabilized at much larger densities, which likely impacted negatively hare, wildcat and domestic cat populations. The domestic cat population was aggregated close to buildings, with a 400 m buffer where vole population is generally lower.
From an applied viewpoint, our results strongly suggest that, in such a highly productive and connective grassland system favourable to grassland voles, any means aiming at increasing the populations of predators during the low density phase (e.g. hedgerow networks, roosts, cats around villages, etc.) should lead to a better control of grassland small mammal populations (slowing down the increase phase) [107]. However, the impacts of a management with large densities of cats around human settlement on other wildlife [108, 109] and pathogen organism transmission (e.g. Toxoplasma gondi) [110, 111] should be considered. Moreover, in such systems and due to unavoidable prey switches some populations like the European hare can be caught in a predation sink and can sustain at low density only. Management options aiming at increasing these vulnerable populations by culling predators (e.g. the red fox, etc.) would conflict with the interests of other stakeholders interested in small mammal pest control. The prohibiting costs and manpower for culling a large number of predators on the long term and the ethical concern of such management should prevent this approach most often shown to be unsuccessful [112, 113] and not accepted socially [29]. Other ways including adaptive hunting plans and demand, modification of habitats and landscapes favouring other equilibriums in the community should be looked for, which implies evidence-based and constructive dialogue about management targets and options between all stakeholders of such socio-ecosystems [114].
Author contributions statement
PG conceived the study with Régis Defaut, designed the sampling plan and the data base. GC organized the transects and road side counts since 2006, AL collected transect and road-side count data, managed the data base since 2014 and georeferenced the observations. PG, AL, MC, GC participated to the road-side counts. EA provided critical insights about cat ecology. PG analyzed the data and wrote the MS. All authors discussed the results and reviewed the MS.
Supporting information captions
S1 kml file. Location of the study area (can be dropped in a Google Earth window or read from a GIS)
S2 Excel file. Road side counts (sheet 1) and list of species observed (sheet 2)
S3 Excel file. Small mammal data
S4 Word file. Detection functions of diurnal road side counts
S5 Word file. Detection functions of nocturnal road side counts
S6 Excel file. Data for computing theoretical daily food intakes
S7 Word file. Variations in biomass by species.
S8 Word file. Variations in theoretical daily food intake by species.
Acknowledgements
This article is dedicated to Régis Defaut, who deceased in 2008 and has been at the initiative and the true core of the ZELAC, a farmer group experimenting on environmentally-friendly control of A. terrestris populations in the study area, including the road side counts and transects used in this study since 1999. Special thanks to Denis Truchetet and Philippe Guillemard, from the Ministry of Agriculture, who have provided support since the beginning, to Jean-Marie Curtil from the Chambre Interdépartementale d’Agriculture Doubs - Territoire de Belfort and Raphaël Bouquet who have provided data on the evolution of grass and milk production, to Clementine Fristch and Francis Raoul for helpful references about food intakes, to the Maison des sciences de I’homme et de l’environnement Claude Nicolas Ledoux, who provided technical assistance to get IGN data, and to the innumerable FREDON technicians, students and volunteers who participated to the road-side counts for their technical assistance. Legal authorization for night road-side counts has been provided by the Direction Départementale des Territoires du Doubs. This research has been carried out in the Zone atelier Arc jurassien https://zaaj.univ-fcomte.fr and received financial support from the Ministry of Agriculture and from the Région de Franche-Comté.
References
- 1.↵
- 2.
- 3.
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.
- 12.
- 13.
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.
- 37.
- 38.
- 39.
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.
- 59.
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵