Impact of American minks on water vole and root vole populations and populations of the aboriginal predators – eaters of these beneficial prey, in particular, the stoat and great grey owl

In one of our study area on the question, which is the Lovat terrain in Paazierre Forestin northern Belarus,  both riparian vole species (the water vole Arvicola terrestris and root vole Microtus oeconomus) have become scarce since the American mink Neovison vison  has established a dense population.

In the years 1997-2003, water voles comprised from 0 to 23.1% (average 10.9%), of the rodent community (estimated by snap-trapping, controlled for the extent of grassy marshes, season and cycling in Microtus voles); and root voles only 0-8.3(2.6)%. To our regret, we have no equivalent snap-trapping data on riparian voles dating from before the American mink arrived. However, counts of water vole latrines were made in both periods, when the American mink was absent in Lovat terrain (1987-1988) and since it became common there (1996-1997 to 2001-2002). Comparing these counts does show the dramatic decline in water voles (12.3-16.1 versus 0.2-3.2 latrines per 100 m2, respectively) that was well correlated with the spread of American mink. Since the decline in riparian voles, other rodent species have apparently become more common in open grassy marshes: bank vole Myodes glareolus – 7.6-62.7%, on average 36.5%;  other species of Microtus voles – 0.7-50.7%, on average 19.6%;  and Apodemus mice – 8.9-30.0%, on average 19.0%.

The question may be raised, why in Lovat terrain the American mink deteriorated the populations of the water vole and root vole, while the European mink Mustela lutreola did not?

American mink

Two distinguishing traits of the American mink determined this ecological situation. First, there is a specific feature in habitat usage by American minks compared to that in European minks (Sidorovich, 2011). By analysing habitat usage in riparian mustelid species in Lovat terrain, polecat’s features in American mink’s behaviour were revealed, at least, in the warm season. If the European mink looked like a stenobiotic predator of aquatic edges mostly at small streams, then in the warm season the American mink largely inhabited grassy marshes located both in valleys and faraway outside valleys, i.e. in isolated remote marshlands and even rather small patches of open grassy marshes. It was typical polecat’s behaviour in the area. This slightly strange habitat usage in the American mink that naturalised in Belarus may be explained by the following. Presumably, it happened because in North America, where the American mink had evolved, there was no any carnivore species with the niche that the polecat had in Europe (Wilson & Mittermeier, 2009). Perhaps, therefore, the American mink evolved in North America by establishing wider ecological niche compared to that of the European mink in Europe. In turn, similar features of feral American minks were registered in naturalised populations of the species in European ecosystems. So, in Lovat terrain American minks following its natural history tended to inhabit all available grassy marshes that were the only habitats for riparian voles – the water vole and root vole.



This slideshow requires JavaScript.

Another distinguishing trait that is important for the subject in question was that the American mink preyed upon rodents more frequently than the European mink and with high preference of the most beneficial vole species –  the water vole and root vole. Moreover, the first years after the American mink arrival in Lovat terrain, the invader mainly took water voles and root voles (Sidorovich, 2011). Then after the crash in water vole and root vole populations the American mink partly substituted the shortage in riparian voles for other species of small rodents. The portion of rodents in the American mink diet did not change markedly, and riparian voles were a considerable part of rodents taken by it from time to time. Thus, the alien predator suppressed continuously the riparian vole populations.

This slideshow requires JavaScript.

Similar destructive influence of the American mink predation on water voles that led to considerable decrease in  population densities and fragmentation of the distributions of the water vole was found earlier in Great Britain (Macdonald & Strachan, 1999; Macdonald et al., 2002). So, these results support all over again our inference that in Lovat terrain the water vole population was deteriorated by American minks.

Similarly to Great Britain, in Lovat terrain and in the whole Belarus the water vole did not disappear entirely. So, another question arises. Where and in which size population fragments the species survived under the conditions of dense population of the American mink? Concerning the root vole, such a question is less interesting, because in central and southern Belarus with larger marshy valleys than those in Paazerje Forest this rodent species went out of suppression by the American mink. There, the root vole was common in the years of population outbreak in Microtus voles.

In 1997-1998 we did much study on the distribution of the water vole in Lovat terrain and other study areas in Belarus and then in 1999-2006 continued some studies on the question more or less. To reveal where water voles can survive in condition of the American mink expansion, we estimated indixes of water vole abundance in habitats differing in size, configuration and dispersion in the study areas. We classified water vole habitat into blocks depending on its area as large (>100m diameter), medium (either 30-100m in diameter or less than 30 m in width if length is longer than 300m) and small (<30m in diameter), and in relation to the distance from the nearest river bank or shore. The habitat patches located within 300m (0-10 m and 11-300 m) of a river bank were classed as adjacent, those at further distances (301-1000 m and >1000 m) as isolated. Also, we estimated the distance between the surveying habitat patch and the nearest large habitat as well as how many large, medium and small habitats there are in the surrounding area of 0.1 km2. We assumed that the above listed parameters as well as share of marshes determine a rate of the American mink predation on water voles and indicate conditions in which water vole population can survive despite of predation of the American mink.

Besides Lovat terrain other study areas (Palata terrain and Krasny Bor terrain in Paazierre Forest in northern Belarus, Almany forest-bog mosaic and Zvanets swamp in southern Belarus, see the area descriptions by Sidorovich (2011)) were chosen for examination on presence of the  water vole, taking into account its requirements  for habitats as well as specific features of the American mink predation on this prey. The following specificity coming from the American mink were taken into account. In the environmental conditions of Belarus winter population density of the American mink was determined by density of river network (Sidorovich, 1997) and during  winter the predator stayed within valleys mainly (Sidorovich & Macdonald, 2001; Sidorovich, 2011). During long-term snowtracking practice and radiotracking study on the species we have never found an American mink that stayed for a long time somewhere outside valleys in winter period. Probably, in the conditions of deep snow cover the American mink could not survive catching small mammals somewhere outside a river valley or shore as this predator species is not so well adapted to do this under snow as the weasel and stoat are. In marshes located outside valleys, the American mink preyed water voles in the warm seasons mainly. So, by choosing the study areas, we used an assumption that water voles are affected by American minks as less as share of marshes is higher and marshes are distributed commonly.

In Lovat terrain, the presence of the American mink positively correlated with the size of wetland, and negatively correlated with the distance between the wetland and the nearest bank or shore (Sidorovich, 2011). Therefore, water voles could survive in numbers in small habitats that were remote from rivers or lakes. Negative correlation between the water vole abundance index and the size of its habitat, and positive correlation between the water vole abundance index and the distance from the nearest bank or shore  were revealed (Sidorovich, 2011). Similar results on the water vole distribution in conditions of dense population of the American mink were obtained by analysing the other characters of its presence such as number of water vole latrines and feeding places (Sidorovich, 2011).

In the sand dune massif with a number of glacial lakes in Palata terrain in Paazierre Forest  in the conditions of very small share of marshes mainly located within river valleys and at shores of glacial lakes  we found few water vole sings and no water voles were captured (Sidorovich, 2011). There water vole sings were found at one large glacial lake only out of eight glacial lakes surveyed.

In southern Belarus (mainly Almany forest-bog mosaic and Zvanets swamp) in conditions of much higher share of marshes (about 10-40 times) and substantially lower density of river network (two times) than in Paazerje Forest we found many water vole sings everywhere in wetlands. Also, water voles were common in snap-trap catches. In comparison with Lovat terrain, in swamped lowlands of southern Belarus we caught significantly more water voles at river bank (on average 0 versus 3.6 inds/100 snap-trap-nights) and in isolated non-valley habitats (on average 1.4 versus 3.8 per 100 snap-trap-nights. In southern Belarus there were not found any significant trends in water vole distribution in depending on size of habitat patch and distance between the habitat patch as well as distance from the nearest edge of aquatic ecosystems.

So, the obtained data suggest that in Belarus after the American mink naturalisation, water voles could survive in numbers only in conditions of lowlands with large marshes. Thereat quite low density of river network in southern Belarus played an essential role, because usually dense river network supported high population density of the American mink. Approximately, if portion of marshes is higher than 10% and density of river network is lower than 0.40 km/km2, water vole population is dense and it does not decline due to the American mink predation. So far as the winter population density of the American mink depended on density of river network and numbers of water voles was determined by abundance of marshes, in such environmental conditions water voles were too numerous to be exterminated by American minks. In turn, American minks were not so common due to fewer winter habitats (e.g. rivers) were available. So, in the swamped lowlands in southern part of Belarus we found huge population of the water vole that was quite undisturbed by the American mink predation.

There was found that the population decline in the big riparian voles, first of all, the water vole had affected considerably the populations of aboriginal predators – eaters of these beneficial prey. Among such predators there are the stoat Mustela erminea, polecat Mustela putorius, eagle owl Bubo bubo, great grey owl Strix nebulosa, short-eared owl  Asio flammeus, greater spotted eagle Aquila clanga and marsh harrier Circus aeruginosus. This negative influence and respective decline in the populations was distinctively pronounced and differently detailed examined.

Taking into account both aspects, the best example is the stoat (Sidorovich, 2011). It was found that in Lovat terrain and Almany forest-bog mosaic stoats were strongly preferred to inhabit wetlands, mainly valley and non-valley open grassy marshes, where they mostly fed on relatively big riparian voles, i.e. the water vole and root vole. Therefore, the decline happened in stoats after the American mink arrival, when riparian voles were strongly affected by the alien predator, was not surprising. At the same time, in swamped lowlands of southern Belarus the stoat survived in numbers. The results of multiannual monitoring of stoat numbers in Lovat terrain, where riparian voles were considerably affected by the American mink predation was following. American mink arrived in the area in late 1988 and since then the number of stoats has decreased with time in Lovat terrain. Taking into account the data on counting of number of track concentrations along 36 km transect in the Lovat terrain in the early winters of 1984-2004, the following stoat population dynamics may be shown. In the winters of 1986–1990, from 8 to 13 stoats were recorded per km2 of the Lovat valley (on average 11.5). This figure dropped to 5-9 stoats/ km2 (average=6.9) in the same habitats in 1993–1995 and, in the winter of 1997–1998, it dropped further to <4 stoats/km2 (on average 1.8). The same trend was observed in the valley marshes of the Lovat upper reashes, where we traced stoat abundance along 16 km transect in early winters of 1986–2003, where stoat numbers decreased about 2.6 fold over the same period (Sidorovich, 2011).


This slideshow requires JavaScript.

Also, the above results were supported by the experience of the local trappers. In the fur-harvest seasons of the 1970s–1980s, before the expansion of the American mink, in Lovat terrain local trappers targeting European mink usually also caught, by chance, about one stoat for 8–10 European mink. In 1990s the ratio of stoats to American mink in the local trap catch was approximately one stoat to 90–120 American mink.

So, our data have documented a decline in numbers and distribution of stoats following the invasion of the American mink in Lovat terrain, at least. The decreased number of stoats has been observed over 14 years already, so is not likely to be a temporary response to a short-term decrease in prey supply. We assume that the cause was the observed reduction in distribution and density of riparian voles (the water vole and root vole) due to excessive predation by the American mink as it became a common naturalised predator. The widely known foraging preference of the stoat is to hunt the largest prey it is able to catch (Day, 1968; Erlinge, 1975, 1981; King & Moors, 1979; Elmeros, 2006; King & Powell, 2007). Hence, the stoat prefers to exploit either the larger of the prey species in a rodent community, such as the water vole and root vole, or populations of Microtus species, especially when they are at high density and biomass. Therefore, when the riparian vole populations in Lovat terrain declined because of an extra predation of the naturalised American mink, the decline in stoats followed.


Another mustelid species which population declined after the American mink expansion was the polecat (Sidorovich & Macdonald, 2001). In Lovat terrain after the invader got high population density the registered decline in polecats was about two fold. Later this species gradually disappeared from forest ecosystems, became rare in valleys and even in rural areas in Lovat terrain. Such strong decline happened in polecats almost everywhere in Belarus. We revealed many different factors that affected the polecat population (Sidorovich, 2011), and presumably the decrease in number and distribution of the riparian voles was one of them. In semi-natural landscape in Belarus a mixture of grassy marshes and black alder swamped woods in valleys of rivers and especially glacial lakes was found to be the main habitat type, where pregnant female polecats gave birth and then raised the litter (23 out of 27 known findings). In normal weather conditions (i.e. without pronounced drought) we gathered 12-49 (totally 156) scats per such polecat family group that stayed in valleys. Although we took relatively small sample size of scats analysed, the gained data suggested a great importance of the riparian voles in feeding kits: the water vole – 12-83(on average 29)%BC; root voles – 0-71(on average 38)%BC. We faced that amphibians were slightly consumed by polecat kits (0-14, on average 8%BC), whereas adult polecats commonly used these largely distributed prey in Belarus (e.g. Sidorovich, 1997). Perhaps, the decline in riparian voles (water voles and additionally root voles in some areas), which were so beneficial for polecat breeding, may be one of the cause of gradual disappearance of the polecat because of the weakened reproduction. American minks densely occupied the main breeding habitats of the polecat and destroyed their prey populations used for raising kits.

Decline in water voles also affected the populations of the eagle owl and great grey owl. Overwintering of these owl species in Naliboki Forest and Paazerje Forest was mainly supported by feeding on water voles and in case of the great grey owl by preying upon root voles, too (Sidoirovich, 2011). Perhaps, therefore, after the American mink naturalisation we found that these owl species were rarely occurred in Paazerje Forest, where populations of both riparian vole species were suppressed by this alien predator.

This slideshow requires JavaScript.

The following data confirm that the owl species became rare there:  Lovat terrain, the eagle owl –  3.1 inds/100 km2 of valley habitats (the same one pair in 1998-2001) and it was nearly absent in forest-bog mosaic; Lovat terrain, the great grey owl –  3.1 inds/100 km2 of valley habitats and 0.3 inds/100 km2 in forest-bog mosaic (two pairs in 1998-2001); Palata terrain – both species were absent (1997-2004); Krasny Bor terrain, the eagle owl – about 2.6 inds/100 km2 of valley habitats (one pair in 1995-1997) and it was nearly absent in forest-bog mosaic; Krasny Bor terrain, the great grey owl – sporadically occurred. Interestingly, that in 1986-1989 in Lovat terrain before the American mink naturalisation and, hence, before the decline in riparian voles both owl species had higher population densities: the eagle owl – minimally 9.2 inds/100 km2 of valley habitats and it was sporadically occurred in forest-bog mosaic; the great grey owl in valleys – minimally 6.1 inds/100 km2, in forest-bog mosaic – 0.5 inds/100 km2.

In swamped lowlands of southern Belarus, where the populations of the riparian voles were not deteriorated, both the eagle owl and great grey owl were more or less common. Around Zvanets swamp in 1998 we censused 17 eagle owls and 25 great grey owls per 100 km2. In Almany forest-bog mosaic in 1999 we censused 44 eagle owls and 16 great grey owls per 100 km2 of valley habitats and 8 eagle owls and 12 great grey owls per 100 km2 of swamp with many forested islands. Additionally, in Naliboki Forest we found strong positive correlation between the parameters indicating number and distribution of eagle owls and water voles (Sidorovich, 2011).

All these results evidently demonstrate pronounced dependence of the eagle owl and great grey owl population densities and distributions on supply with riparian voles especially with water voles and, in turn, suggest the considerable decline of the owl species populations over the American mink expansion and naturalisation.

One more feature is worthwhile to notice. From the above-presented data on the great grey owl distribution it appears that in Belarus this boreal owl species was much commoner in south-western part than in north-eastern part of the territory, while it would be more natural for the great grey owl to be commoner in more northern latitudes. There might be some secondary impact causing such a strange phenomenon. There is much support that it resulted from the decline in riparian voles affected by predation of the American mink as an extra and numerous predator species. As you found above, the decline in riparian voles was much stronger in the northern part of Belarus.

Concerning the short-eared owl, in 1986-1989 it was registered as quite common species in a marshy part of the Lovat valley before the American mink arrival – about 246 inds/100 km2 of the valley, whereas afterwards the owl species got fairly rare there –  about 15.4 inds/100 km2 of the valley in 1997–1999. In 1997 and 1998 in Zvanets swamp in southern Belarus we counted about one pair per one km2, i.e. 200 inds/100 km2. We did not obtain muchl data on the diet of this rodent eater, but again it is likely to accept the same hypothesis related to riparian voles that could explain the results on the short-eared owl.

Short-eared owl
Short-eared owl

Another predator species that was presumably suppressed by the decline in riparian voles was the greater spotted eagle. In 1995-2003 in Lovat terrain (i.e. after the American mink naturalisation) this diurnal raptor was quite rarely occurred – actually one breeding pair in marshy valley of Lovat river for the majority of the years that gave average density about 5.4 inds/100 km2 of valley habitats. Although the water vole was rare in occurrence, the greater spotted eagles consumed the prey in considerable portion – 18.7%BC (Sidorovich, 2011). On the other hand, in Almany terrain, where riparian voles survived in numbers, there was found relatively dense population of this raptor (e.g. Dombrovski & Ivanovskij, 2005).

So, in Belarus there was registered detrimentally destructive effect of the American mink on the populations of the riparian voles, first of all, on the water vole population. This resulted dramatic deterioration in the predator guild – eaters of these beneficially big prey.