Weasels In Temperate Farmland And Forest

Compared with the far north, most temperate habitats support a much wider variety of prey and a larger community of predators, many of them resident all year round. Snow cover may be short-lived, and weasels must not only share the small rodents with larger predators, but they also must also watch out for themselves (Chapter 11).

Vole populations in many temperate habitats fluctuate annually, increasing during breeding seasons and decreasing between breeding seasons. The swings from low to high and back seldom reach the spectacular amplitudes typical of the far north, but fluctuations in vole populations are often large enough to make a lot of difference to the hunting prospects of individual weasels. In Britain, for example, in the deciduous forest of Wytham Wood near Oxford, the combined density of wood mice and bank voles remained between roughly 10 and 30 rodents/ha throughout the 1960s and 1970s (Southern & Lowe 1982). During those years, tawny owls and weasels were the two most important vertebrate predators in the forest. In 1968-1969, the known resident weasels ate on average about 8% to 10% per month (range 2% to 20%) of each of the populations of bank voles and wood mice (King 1980b). These losses accounted for only a small proportion of the 12% to 55% of voles and mice disappearing each month, and for perhaps 14% of the total production of rodents in the forest (Hayward & Phillipson 1979). Predation by weasels also had little effect on the average survival time ofindividually marked mice (Flowerdew 1972, pers. comm.). On the other hand, the tawny owls were also taking a regular toll, and at times it was substantially heavier than the weasels took. Tawny owls removed on average 15% to 33% of the mice and voles each 2 months.

These two sets of results from Wytham referred to short periods that did not coincide, but they confirmed that common weasels and tawny owls were the two most important predators of voles and mice in that woodland (Southern & Lowe 1982). Years later, the same conclusion was reached after a much more comprehensive and thorough study in Bialowieza Forest in eastern Poland. There, a suite of 23 species of predators (of which the most abundant were tawny owls, common weasels, buzzards, and pine martens, in that order) hunted bank voles and yellow-necked mice all year round. Four species (tawny owls, common weasels, pine martens, and red foxes) accounted for, on average over three successive winters, virtually all (>90%) of the toll on both bank voles and yellow-necked mice (Table 7.3); stoats added another 1% to 2% on each. The total number of rodents removed (28 to 35 voles and 14 to 17 mice per ha over 197 days—about 1 kg of rodent meat per ha) was about the same as the total decrease in rodents over the winter: Voles declined from 35 to 8 per ha between autumn and spring, and mice declined from 24 to 3 per ha (Jgdrzejwski & Jgdrzejwska 1993).

Yet, all that concentrated predatory power could not prevent the Polish rodent populations from irrupting when the oak, hornbeam, and maple trees produced a heavy crop of tree seeds. In New Zealand, Blackwell et al. (2001, 2003) and Ruscoe et al. (2003) likewise concluded, from independent data sets, that predation by stoats is not likely to prevent an irruption of feral house mice after a heavy seedfall in southern beech forests, even though predation may have a significant effect on mouse populations between irruptions (Choquenot & Ruscoe 2000).

By contrast, in southern Sweden the local populations of common voles and wood mice vary through the year, but hardly at all from one year to the next (Erlinge et al. 1983, 1984). On a 4-km2 study area of meadows and grazed pasture near Lund, common voles ranged only from 8 to 10 per ha in May and June to about 50 per ha in August and September. Erlinge and his colleagues set out to document the effects of the entire community of predators on the small mammals. They estimated that, in 1975 and 1976, the number of rodents produced roughly equaled the number eaten each year by the generalist predators (fox, feral cat, badger, buzzard, and tawny owl) plus the number eaten by the predators that specialize on rodents (stoat, kestrel, and long-eared owl). Even

Table 7.3 Impact of Predators on Bank Voles and Yellow-Necked Mice in Bialowieza

National Park

Number of rodents removed from 1 ha in 197 days in autumn and winter

Table 7.3 Impact of Predators on Bank Voles and Yellow-Necked Mice in Bialowieza

National Park

Number of rodents removed from 1 ha in 197 days in autumn and winter

Species

1986-

1987

1987-

1988

1988-

1989

n

%

n

%

n

%

Voles

Mice

Voles

Mice

Voles

Mice

Voles

Mice

Voles

Mice

Voles

Mice

Tawny owl

16.3

11.7

58

70

19.3

11.1

56

71

22.3

9.8

65

70

Common

4.1

2.3

15

14

7.3

2.2

21

14

5.1

2.0

15

14

weasel

Pine

3.6

1.6

13

10

5.1

1.5

14

10

4.9

1.6

14

11

marten

Red fox

2.6

0.6

9

4

1.4

0.4

4

3

0.7

0.2

2

1

Stoat

0.5

0.1

2

<1

0.5

0.1

1

<1

0.5

0.1

1

1

(From J^drzejwski & J^drzejwski 1993.)

(From J^drzejwski & J^drzejwski 1993.)

though rodents constituted only about 15% of their diets, the generalist predators accounted for over 75% of the rodent losses (Table 7.4). The stoats' share in the total was under 10%.

The key conclusion from Erlinge's wide-ranging study, the first to give us a clear picture of predation processes at the community level, is that communitylevel processes are important. The data allowed the losses due to the combined force of predators to be set against the production of all the small mammals together. The losses to the rodents were especially heavy in spring, when other favored prey, such as young rabbits, were not available. The voles started their breeding season surrounded by persistent predators poised to snap up the young as soon as they emerged from their nests. This heavy predation delayed and reduced the rodents' recovery from the winter nonbreeding period.

Predation was also particularly heavy in autumn, because of the rapid functional response of the generalist predators to the large increase in numbers of rodents through the breeding season. The net result was that rodents could never escape the attention of predators: The generalists were always ready and waiting, and the specialists joined in when they could. Stoats in this situation were almost as much at risk as their prey, because, when the population of rabbits declined, the stoats were forced into severe competition with the generalist predators for the remaining rodents (Chapter 10).

Populations of voles on meadows and arable land are capable of larger annual fluctuations than those of woodland species, but predators can still account for most new rodents produced. The total, annual production of field voles in England is estimated to be between 677,000 and 982,000 (Dyczkowski & Yalden 1998). The total annual consumption by all predators combined is estimated to be roughly 980,000 voles, of which two specialist predators on voles, common weasels and European kestrels, plus two generalist predators, red foxes and feral cats, account for 85%. Predators of 10 additional species divide the other 15% between them. Common weasels are estimated to kill some 22% of the voles lost to predators, and stoats 4%. During most years, the total production of young voles is roughly matched by total predation, explaining the relatively small changes in the field vole populations from year to year in these habitats.

Table 7.4 Annual Production and Mortality of Field Voles and Wood Mice on 4,000 ha of Marshy Meadows and Pasture in Southern Sweden

No. eaten per year by predators

No. produced

Total Generalists Specialists

% of total eaten by stoats

Voles 171,400

156,865 120,700 36,165

9%

Mice 20,100

21,546 17,180 4,366

7%

These ball-park figures necessarily conceal some important local and seasonal variations. For example, in one study in Kielder Forest, common weasels in overgrown clearcuts were seldom numerous enough to account for more than 5% of the variation in vole survival, reaching 20% only for a few months in summer when the weasels were most abundant (Graham 2002).

Goszczynski (1977) described how a team of Polish workers, interested in the causes of the population fluctuations of the common vole that periodically damage agriculture, attempted to account for all predation processes over 3 km2 of mixed farmland. They examined the diets of martens, foxes, badgers, feral cats, and four species of hawks and owls as well as of common weasels. The study ran for 3 years, and covered a complete fluctuation in numbers of voles, from a low in late 1970 through the peak in 1971 (>330 voles per ha) to the next low in 1973. The proportion of voles removed by all the predators combined was high to start with, when the voles were scarce, and about three quarters of the total mortality of the voles at that time was due to predation. But by the time the voles had reached their peak, they were so abundant that predation could account for very few of their numbers and less than half of their mortality, even though by that time all the predators were living almost entirely on the voles, and some of them, including the common weasels, had also increased in numbers.

On the other hand, as the vole numbers declined, the predators were still numerous, so the ratio of numbers ceased to favor the voles. The increasingly desperate predators searched out almost every single vole that was left, reducing the population to about 1 vole per ha. At the height of the slaughter, predation removed an estimated 31% of the energy available to predators in the vole population. The common weasels' share was calculated at 11%, which put them in third place after foxes (37%) and feral cats (29%) in the number of common voles eaten. Unfortunately, the team's density estimate of one weasel per 3.9 to 4.5 km2 was probably an underestimate (see Table 10.2). Our guess is that the real contribution made by common weasels to the toll was much higher, especially when the vole population was decreasing.

On North Farm, a game estate in Sussex, England, Tapper (1979) followed the changes in the numbers of field voles from 1971 to 1976. The 2.5-km2 study area of gently rolling chalk downland was mostly divided into huge arable or pasture fields with patches of woodland and rough grass. Field voles and common weasels both avoided the open fields, which were frequently rolled, mowed, or heavily grazed, so both lived together in the few undisturbed areas and could be censused there easily. The voles were declining at the beginning of the study, fell to very low numbers (about 20 voles per ha) in 1973, shot up to around 300 voles per ha in 1974-1975, and declined again in 1976. The numbers of weasels caught followed the numbers of voles, but lagged behind by about 9 months (see Figure 10.7). The weasels ate about three to four times more voles per head in the years when voles were most numerous (54% of their diet in 1975 compared with 16% in 1973 when voles were few). Weasels also doubled their own numbers, thereby increasing the number of voles they could remove almost 10-fold.

Wood mice are usually much less common than voles, and also weasels find them more difficult to kill (Chapter 6), so predators hunting mixed populations of voles and wood mice tend to take fewer mice than voles (Table 7.3). The mixture of the two types of prey has consequences for the weasels. On a 150 ha area of farmland in France, the weasels observed by Delattre (1984) removed a substantial proportion (16% of total rodent biomass per month) of a mixed population of rodents when it was dominated by voles, but when mice predominated, both the weasels and their impact declined (Table 7.5).

Severe predatory impacts by weasels in temperate countries are seen only on the rare occasions when a population of rodents is confined to a limited area. In the days before combine harvesting, field crops were stacked into ricks in late summer, and dismantled only in autumn or winter when farm workers had time to do the threshing. A large, well-built grain-stack was a least weasel's idea of paradise: warm, dry, safe from large predators, and overflowing with thousands of rodents. (Modern hay-barns are good, too, but without the concentrated supplies of grain stored in the old grain-ricks, they are not quite of the same class.)

Farm workers welcomed weasels to a rick, with good reason. Over the winter of 1948-1949, least weasels occupied fully 90% of the ricks examined in the Miknov district, near Moscow, by Rubina (1960). In nearly all occupied ricks, there were fewer rodents than in the few ricks not visited by weasels. A more recent example is the study of the overwintering habits of radio-tagged meadow voles in New York (Madison 1984) that was wrecked by a stoat who entered the enclosure. The stoat killed more than half the tagged voles and took over a nest, lining it with fur and carrying back to it the voles it had killed, complete with their transmitters.

Predation by weasels in temperate countries affects not only small mammals. Wytham Wood in England has been the site of more than 50 years of research on the population dynamics of tits (Paridae). These small birds nest in

Table 7.5 Impact of Common Weasels on a Mixed Population of Common Voles, Field Voles, Bank Voles, Wood Mice, and Yellow-Necked Mice on 150 ha of Farmland in France

Spring 1978

Spring 1979

Spring 1980

Combined density/ha

7.2

6.2

6.8

Proportion mice/voles

41/59

81/19

80/20

Number of weasels

Male

9

5

2

Female (pregnant)

6 (all)

2 (1)

0

% biomass rodents taken by weasels

16

10

3

(From Delattre 1984.)

(From Delattre 1984.)

natural holes in trees, but also readily use the 1,000 artificial wooden nest boxes provided by researchers. Annual records of nesting success include data on which nests are destroyed by predators, most often by common weasels (Figure 7.1).

For the first 10 years of monitoring, until 1957, weasels raided only 0% to 8% of the nest boxes per season. Then, suddenly, up to half of the boxes were raided in each of the next few years. Dunn (1977) analyzed the detailed records available and showed that year-to-year variations in two factors alone were sufficient to regulate the population density of the tits: clutch size, which was determined largely by food supplies, and hatching success, which was strongly influenced by weasel predation. In turn, the extent of predation in any one year was affected mainly by whether small rodents in the wood were at low density during the nesting season (see Figure 13.3). The weasels sometimes also managed to catch the female bird on the nest (each nest box has only one entrance, and only females brood the eggs), and this might explain why female tits live slightly shorter lives than males.

Figure 7.1 A common weasel in the act of raiding a nest box of a great tit and removing a chick. (Redrawn from a photograph by C. M. Perrins, published by Dunn 1977.)

One might ask why the effect of weasel predation on the tit population was apparently greater than on rodents in the same habitat. There are several possible answers: The nest boxes were conspicuous and easily found; the nests were available only over a short season; lost clutches were rarely replaced more than once; on average there were fewer nests than rodents per weasel (nests about 2 per ha; rodents about 21 per ha); and weasels were by far the most significant predator on tit nests, whereas tawny owls were probably more significant than weasels as predators of rodents.

Since 1976 the problem for the tits (and for the researchers working on them) has been resolved by the installation of concrete, weasel-proof nest-boxes (McCleery et al. 1996). After that, nest predation dropped from 30% to less than 5%, the mean number of fledgelings per nest rose from 3.4 to 6.3, and the chances of adults living to the ripe old age (for a small bird) of 5 years was greatly increased. The weasels no doubt continue to raid the unprotected nests of other birds, as they do all over the world (p. 307).

Was this article helpful?

0 0

Post a comment