Snow provides opportunities to collect data that cannot be collected during warm weather. Weasels travel across the snow with a half bound, the front paws landing nearly simultaneously but with one slightly in front of the other. The hind paws land in the exact prints of the forepaws. Prints of the paired paws may be as close together as 10 to 12 cm for a least weasel, but can reach over a meter apart when a large weasel runs down hill. The tracks of weasels and all their close relatives (ferrets and polecats, minks, martens, wolverines) are the same except for size, and are unmistakable once learned (see Figure 6.1).
The ancient art of reading snow tracks has been developed into a standardized density index, first by Russian game biologists and fur trappers (Aspisov & Popov 1940; Teplov 1952), and more recently by the Scandinavian school of ecologists. For example, on a 28-km2 study area at Alajoki, western Finland, E. Korpimaki and K. Norrdahl defined a series of fixed transects across farmland and forest, and checked them systematically for weasel tracks after every snowfall (Korpimaki & Norrdahl 1989a; Korpimaki et al. 1991).
To estimate the density of least weasels, the transects were 60 m apart, equivalent to the mean home range width of both male and female least weasels according to Nyholm (1959b). They measured the footprints to distinguish the sex of each individual (prints averaged 4.6 cm for males and 3.1 cm for females; stride lengths 56.3 cm and 30.2 cm, respectively [Nyholm 1959b]). Only a few females were recorded (the pooled ratio of females to males over the four winters 1984-1987 was 6:30), but that was not surprising since females spend much more time in under-snow burrows (Chapter 2). They assumed a 1:1 sex ratio, and calculated density indices by doubling the number of males, but without correcting for the density of voles. They came up with figures ranging from 2.4 to 13.0 least weasels per km2, depending on the density of the vole population (Korpimaki & Norrdahl 1989a:209).
Variations on this method are still widely used in Finland (Aunapuu & Oksanen 2003), but there are three potential problems. The first is that weasels commonly dive down into or pop straight up through the snow blanket. Although it might look reasonable to assume that a track coming up from under the snow was left by the same weasel that went down nearby, that is not always true. Home ranges of males and females overlap, so it is easy to confuse individuals.
The second is that snow conditions affect both the numbers of tracks recorded and apparent sizes of individual footprints. Jgdrzejwski et al.(1995) showed that the depth of the snow explained nearly 40% of the variation in numbers of tracks of common weasels counted per km per day, because when snow was more than 40 cm deep the weasels spent most of their time in the subnivean spaces. And, after long experience of tracking mustelids in the snow, R.A. Powell (unpubl.) has records showing that different snow conditions can change the apparent track size observed for the same individual on the same day.
The third problem is that the paw measurements of the two sexes of all mustelids we know overlap to some degree, even for those species with large sexual dimorphism. These three problems combine to make counting individuals and identifying their sex from tracks in the snow very imprecise.
In countries and seasons where snow does not provide a ready-made tracking medium in which to record footprints, it is necessary to provide one. Fortunately, the perpetual curiosity of weasels makes it possible to design a simple and effective method. Artificial tunnels containing tracking papers and ink pads or sooted surfaces (as described in Chapter 8) are set out in grids or lines, and have two main uses. When tracking tunnels were used in remote places as a simple presence/absence survey tool, a durable ink (King & Edgar 1977) was advantageous because it extended the possible period between checks. Alternatively, if routine population monitoring is required, a rough density index can be calculated from the proportion of tunnels recording weasel tracks over a given short period.
Population indices are more accurate if the papers are changed frequently. The longer the period is between checks, the more tracks can appear on the paper, to the point where identification sometimes become difficult. Moreover, if more than one weasel track appears on the paper, there is no way to tell whether they were made by one weasel visiting more than once or more than one weasel. These problems can be reduced by checking the tunnels more often, or reducing the sample period, especially as short-lived, simple dyes such as food coloring can then be used for the "ink."
Tracking tunnels are especially useful because they offer a cheap and easy way to answer one of the most important and difficult questions about any species of weasel: Where are they? Or conversely, are there any here? Students planning to begin a radio-tracking study of weasels, or conservation managers needing to protect especially valued breeding birds from nest predation, need to know where weasels are and how many of them are there.
Tracking tunnels are a simpler method of finding this out than are traps, especially if the weasels are not to be removed when found. Tracking tunnels do not interfere with weasels' normal movements and can be set in arrays of hundreds at a time, and the prints are usually clear and easily measured. The question is, are they reliable enough to be a valuable tool for research and management?
The answer seems to be, as so often, yes and no. There is a correlation between tracking rates and weasel density, but it is not linear. In Britain, Graham (2002) used unbaited tracking tunnels, calibrated against live trapping, to show that the correlation between tracking rates and number of weasels live trapped was close but varied with vole density and with season. The same number of tunnels with common weasel footprints represented a higher density of weasels when voles were abundant than when they were scarce, and more in summer and autumn than in winter and spring. So long as these complications are taken into account, Graham (2002) concluded, tunnel tracking is a reliable means of assessing the abundance of weasels.
In New Zealand, tracking tunnels are now widely used, usually baited (Clap-perton et al. 1999), to index population densities of rodents and stoats (Brown et al. 1996; Brown & Miller 1998). Tracking tunnels can document population changes, for example, the reduction in numbers of stoats after a control operation (Murphy et al. 1998, 1999), but less reliably in autumn and spring than at other times of year. When the immigration rate is high, such as in the autumn after a good season for producing young, the residents removed might be replaced too quickly for tracking indices to detect the difference (Dilks & Lawrence 2000).
The ultimate challenge in weasel spotting is to detect the arrival of a single colonizing individual in a large area devoid of them, for example, after a lemming crash in the far north. There, one can exploit a simple index of mustelid activity, such as snow tracking or the spring distribution of lemming nests raided by stoats over winter, to detect as few as two stoats in 1,000 ha (Sittler 1995). Such options, however, are open only to a few. For the rest of us, tracking tunnels are the next best thing.
To see why, consider the model constructed by Choquenot at al. (2001). This model predicted that about 350 tunnels would be needed to detect the presence of a single stoat on an area of 10,000 ha (100 km2) with 75% confidence (assuming each tunnel samples 1.5 ha, a probability of 0.7 that a stoat encountering a tunnel will enter it, and that home ranges average 50 ha with 20% overlap). If five stoats were present, only about 50 tunnels would be needed to detect at least one of them with the same confidence. Fifty live traps on 100 km2 (one per 200 ha) would be extremely laborious to operate, assuming they would have to be inspected daily: 350 would be impossible. Conversely, in an area supporting an established population of weasels at a density of say 10 per km2 (Table 10.2), that is, 1,000 individuals in 100 km2, the probability of a very small array of tracking tunnels (<50) detecting at least one weasel rises to 99%.
Was this article helpful?