Of Live Populations

Estimates of absolute density of weasels are rare, because the statistical techniques for calculating the numbers of live animals per unit area impose important conditions on the field data that are hard to meet, including the following:

1. The population must be counted without removing the animals or frightening them so much that they refuse to be counted again. The best available, though not ideal, method for weasels is still the old one of using live traps to capture, mark, and recapture repeatedly the resident individuals of a given area. But confinement in a live trap is probably unpleasant for weasels, and it tends to discourage at least some of them from risking such an unhappy experience again (King et al. 2003a). Others become "trap happy," especially if wild prey are so scarce that a night in the trap becomes a price worth paying for a free meal.

2. It must be possible to define the area sampled, so as to calculate the catch per unit area, and that is not only very difficult but also has a disproportionately large effect on the results if the boundaries are misjudged.

3. It must be possible to estimate the numbers of animals entering and leaving the population, and the number that are present but not caught. Not every resident will be caught during every trapping session, but if it is caught again in a later session it is usual to assume it was there all along, even though this sometimes seems quite unlikely. For example, one radio-collared common weasel was neither recaptured nor detected by radio for 14 months, until it was again recaptured not far from the site where it was first collared (Delattre et al. 1985)

4. Some individuals will not enter traps at all, so every study of weasels should allow for some unspecified number of shadowy, unidentified figures in the background. The proportion of these probably increases when food is abundant (Teplov 1952; Alterio et al. 1999; King & White 2004).

The problems do not stop once these hurdles are crossed and a set of field data is safely entered into the computer. The simplest way to estimate the population density is the "calendar of captures" or "minimum number alive" method, which compiles regular totals of the number of residents known or assumed to be present each session. Unfortunately, live trapping is very labor intensive, so for very mobile animals such as weasels it is usually impossible to observe more than a few individuals at a time, or to define the boundaries of the sample area. This means that the population estimates are so much affected by statistical errors that they may be quite wildly wrong. More sophisticated methods designed to overcome this problem (e.g., Efford 2004) usually demand far more data than a short-term live-trapping study on weasels can supply.

A few outstanding studies have met this problem by deploying a team of field workers over a large area for several years. The first of these was led by Sam Erlinge (1983), whose group documented the population dynamics and predator-prey relationships of stoats on a 40-km2 area of farmland and marshes in southern Sweden over 5 years. Erlinge was the first to test the accuracy of live trap-night indices (live captures per 100 trap nights) on a population of stoats of known density, and he concluded that they are a reasonably reliable guide to changes in relative numbers.

Another team study was done by Jgdrzejwski et al. (1995) on a 47.5-km2 area of the Bialowieza National Park, eastern Poland. The Polish group integrated data from grid-based snow tracking, live trapping, radio tracking, and computer analyses to convert live trap-night density indices into absolute numbers of common weasels with remarkable precision (r = 0.9, P = 0.002) (Figure 10.1).

One possible way to get density estimates for larger areas is to take reasonably accurate measurements of density in small areas and scale them up. That approach can lead to completely unrealistic results if done uncritically, because it assumes that the whole of the larger area is covered by the same habitat sampled in the smaller one. On the other hand, it is possible to derive a rough estimate of the total population of a large area by multiplying the typical density of a species in different habitats by the area of those habitats available.

Stephen Harris and his team (1995) were the first to attempt a marriage between habitat distribution surveys and habitat-specific density indices, in an attempt to calculate the total national populations of all 63 species of native and introduced mammals in mainland Britain. National biodiversity strategies being developed by countries that are signatories to the agreement made at the United Nations conference at Rio de Janeiro in 1992 need this sort of information, however rough it may be. Under European Union legislation, the monitoring of endangered wild mammal populations is now a statutory responsibility, so the UK Joint Nature Conservation Committee commissioned r\i r\i

Figure 10.1 The correlation between the absolute density of common weasels and a density index in Bialowieza Forest, Poland. (Redrawn from Jgdrzejewski et al. 1995.)

Index - Weasels /100 TN

Figure 10.1 The correlation between the absolute density of common weasels and a density index in Bialowieza Forest, Poland. (Redrawn from Jgdrzejewski et al. 1995.)

Harris and his team to identify which species could be of conservation concern in Britain.

The easy part was to list, from existing survey data, the total national areas of various defined habitats. The hard part was the next step, to get from known figures on the distribution or home ranges of weasels in particular places to a general figure for density per square kilometer of the same habitat. The problem is that home range data can be very misleading if they include, as they usually do (Chapter 8), a lot of spatial overlap and temporal variation. Eventually Harris et al. simply decided to assume the rather generous figures of six stoats per km2 in all types of woodland, and one to two per km2 for various types of grassland, which produced a calculated prebreeding national total for Britain of 462,000 (about two per km2 overall). They calculated the numbers of common weasels from the ratio of common weasels to stoats in gamekeepers' bags, and came up with a national total for them of 450,000.

There is no way to guess how close these estimates are to the mark, but from 1970 to 1995, British gamekeepers working on habitat favorable for wildlife killed about 1.5 to 2.0 stoats per km2 every year (Tapper 1999), so the prebreeding population must be able to maintain a steady harvest at that level, at least in those habitats. The recent drop in gamekeepers' tallies of stoats and common weasels in Britain is probably due to a decrease in trapping effort, not in the national populations of stoats and common weasels (McDonald & Harris 1999).

In the coastal dunelands of the Netherlands, the prebreeding density of stoats through the 1950s and 1960s was about two to three per km2 in early spring and six to seven per km2 in summer; the annual harvest was about four per km2 (Mulder 1990). In Fennoscandia, numbers of least weasels are reckoned to range from one to 20 per km2 and stoats from 0.5 to 2.0 per km2 (Hanski et al. 2001).

Calculating a broad-scale average density for American weasels (all species) would be more difficult, and the only estimate we have found is that given by Craighead and Craighead (1956) for lower Michigan in 1942. They were not specifically studying weasels, which are hard enough to find at any time, so they probably underestimated the numbers of weasels in their study areas. From tracks, observations, and live trapping, rather than home range data, their figures of 27 to 36 per township (36 square miles, 93 km2) convert to 0.29 to 0.38 per km2.

In New Zealand, knowledge of the range of absolute densities of stoats before and after control operations is critical to developing effective management programs to save the national icon, the brown kiwi, from extinction. The prospect of losing the kiwi has spurred intense interest in research on the biology and management of stoats there, described in Chapter 13. Independent calculations of stoat densities in New Zealand are few but fall roughly into the same range as for stoats in Britain and the Netherlands.

In a southern beech forest after a productive summer when mice were abundant, Murphy and Dowding (1995) found eight stoats (four male, four female)

resident on 150 ha (five per km2) between January and May 1991. In similar conditions but a different place, Alterio et al. (1999) calculated that there could have been a postbreeding population of three to seven (average 4.2) stoats per km2, decreasing to two to four per km2 (average 2.5) nonbreeding adults in the following winter. After a good summer for mice, densities of stoats can remain higher than normal for another 6 to 12 months in the absence of control measures (Figure 10.4).

Two quite different methods of estimating absolute density from different beech forests have been applied in New Zealand recently. One uses hair tubes (Chapter 8) baited with rabbit meat to collect samples of hairs from stoats, from which individual DNA profiles can be extracted (Gleeson et al. 2003). In the first trial, 60 hair samples were collected each week during a month-long trial, from plastic tunnels placed 250 m apart on a 3x3 km grid (9 km2). DNA profiles were obtained from about 80% of the samples, and from them 30 different stoats were detected (3.3 stoats per km2). A second method uses total removal trapping to calculate absolute density of an undisturbed population. On Te Kakahu (Chalky Island), off the southwest coast of Fiordland, the first forested island where this was tried (Chapter 14 and see Willans 2000), 16 stoats were taken from 514 ha (3.1 per km2) within a few weeks in 1999. Anchor Island (1,100 ha excluding a 200-ha lake) was cleared of19 stoats in winter 2001 (1.7 per km2) by essentially the same method (M. Willans, unpubl.). On a 750-ha peninsula in a lake, McLennan and his team (McLennan et al. 1996; McLennan, unpubl.) removed 65 stoats in 3 months during a mouse peak year (nine per km2), but many fewer per 3 months (zero to two per km2) in mouse-poor years. Stoats are good swimmers, so the water protecting the peninsula on three sides would only have slowed down immigration, not prevented it.

How many stoats live in New Zealand? Summer irruptions of stoats in beech forests are short-lived, but recur every 3 to 4 years when mice are abundant after a seedfall. In other types of native forest, where absolute stoat densities have never been measured, relative density indices average around the lower end of the range for beech forest (King et al. 1996). The combined area of native forest patches of all types in New Zealand was 62,800 km2 in 1993 (Taylor & Smith 1997). If the general prebreeding density of stoats in native forest is two per km2, then the total forest population in spring could be about 125,000. Other habitats that might be occupied by stoats, such as exotic forest (14,000 km2), crops (4,800 km2), and tussock grasslands and pastures (135,200 km2) cover another 154,000 km2 of New Zealand. The density of stoats in them is unknown but probably low, especially in open country (Keedwell & Brown 2001), but even if it averaged only 0.5 per km2, that still makes another 77,000 stoats.

Obviously, it is impossible to estimate the total population of stoats in New Zealand with any confidence. At first glance these estimates look wrong, because they are a lot lower than given above for Britain, a country of roughly the same size. The data for New Zealand are few, but the figures we do have suggest lower average prebreeding densities of stoats in all New Zealand habitats, especially in the much larger total area of forest, than were used for the British estimate. Nevertheless, New Zealand could easily support 200,000 of these easy-to-hide little predators in the spring of a normal year, and for a short period in summer after an especially good breeding season, many more. Half a million stoats could do a lot of damage in a very short time. It is no wonder the New Zealand conservation authorities worry (Chapter 13).

Was this article helpful?

0 0

Post a comment