Key Words : Alpine environment, Efficiency, Ethical aspects, Marmots, Marmota marmota, Population monitoring, Trapping methods.

The will of humans to preserve the environmental biodiversity (United Nations 1992) and to manage wildlife as natural renewable resources leads to necessary manipulations of wild animals for re-enforcement as well as for re-introduction as indicated by IUCN (1987). Live trapping is then a major and unavoidable way to manage wildlife (Flowerdew 1976). Improving these methods is needed for different reasons such as ethics, socio-economics, efficiency and feasibility. As far as the law includes necessary welfare in animal use by people we have to promote the less injurious ways of trapping as well as to minimize the stress induced by catching.

Marmots (genus Marmota, fam. Sciuridae, Marmotini, Rodentia) are medium-sized (adult body mass usually from 4 to 6 kg), diurnal, sedentary and social mammals (except the woodchuck M. monax). They shelter in burrows dug in open meadows and grasslands, both in plain and mountain (up to 4000 m a.s.l.), from 35° North to 72°N (Barash 1959).

Wildlife management is a matter of concern for some species or subspecies of marmots in the world. Two species are quoted in IUCN Red List (1990): Marmota menzbieri, vulnerable in Kazakhstan and Uzbekistan, and M. vancouverensis, endangered in Vancouver Island (Johnson 1989). In Yakutia (Russia) M. camtschatica and M. sibirica in Mongolia are heavily exploited for fur, and these species are consequently endangered (Orlov 1989). The Carpathian subspecies of M. marmota (M. m. latirostris) is quite extinct in Polish Tatra Mountains and reaches a very low level in Slovakian Tatra Mountains (Pucek 1989). This subspecies disappeared from Romania at the end of 19th century (Almasan 1981) and was replaced unfortunately by the Alpine subspecies (M. m. marmota) (Ramousse and Le Berre 1993). For these reasons, a good knowledge of safe trapping methods is of great importance in management and conservation of marmot biodiversity.

Marmot live trapping is performed anywhere in the world for scientific purposes (Armitage 1982; Lenti Boero et al. 1988; Arnold 1990; Bassano et al. 1993; Perrin et al. 1993; Sala et al. 1993), health monitoring (Tleuzab’lov et al. 1961; Bibikov et al. 1973; Pole 1992), management purposes or wildlife conservation (Bibikov 1989; Ramousse et al., 1992).

The aim of this paper was to strike the balance between different available methods for trapping live marmots in a long term population study. General works on trapping methods have been reviewed (Chaigneau 1968; Giles 1969; Young 1975). Different authors have shown that mammal species have different trappability with different kinds of traps (Chitty and Kempson 1949; Beacham and Krebs 1980; Jensen et al. 1993). It led us to compare 4 kinds of live traps: 3 among the most often quoted in literature for trapping marmots : one-door and two-doors live traps, jaw traps and snares, and one commonly used in France, jaw traps. The criteria for comparison were : efficiency (rate of capture/day/trap), time and environmental effects, selectivity (did a particular trap introduce a bias either for age or sex ratio of caught animals), ethical requirements (animal welfare, recapture ability, probability of survival of released animals), handiness (size, shape, weigh, positioning), and cost.

For technical reasons the experimentation has been performed over two summer seasons (1992 and 1993). Live traps and jaw traps were tested during the first year and live traps and snares during the second year.

The experiment was performed in the central area of the Ecrins National Park on the site of Plateau de Charnières (44°66 N, 4°05 E, Central Alps, France) around 1600 m a.s.l. The plateau extends on 50 hectares and is surrounded by steep slopes both northwards (sunny side) and southwards (shady side). The choice of this site was related to its early accessibility in spring. The fact that this protected place was yet used by farmers (hay cutting) and frequently visited by hikers played also a role as such places need accurate conservation strategies to preserve the natural and cultural heritage represented by reap-meadows in a subalpine environment (Cortot et al. 1992).

In this experiment four kinds of traps have been tested. Ten traps of each kind were used at the same time.

One-door live traps (1DT) : This pattern was used by our research team for catching marmots since 1990 (Perrin et al. 1993). Its size was 75 x 25 x 25 cm, and its weight 3.5 kg. The door was closed by a treadle with an outside steering-gear.

Two-doors live traps (2DT) : It was a new pattern improved from the former (Le berre et al. 1993). Its size was 100 cm x 25 x 25, and its weight 5.0 kg. The doors were closed by acting a treadle with an outside steering-gear.

Both live traps were made of wire-netting of 2.4 mm in diameter with welded rectangular meshes (50 x 25 mm). Particular care was made to the locking of the doors as their joint may be a frequent cause of escaping after capture (Ludwig and Davis 1975). Both traps were placed at 1 m from the burrow opening, the entrance directed towards the opening. Traps were baited with salt stone or apple. When a marmot was caught, the trap was covered by a blanket in order to quiet the animal in dark. The trap door was surrounded by a linen bag (mail-bag) and opened, preventing any hand contact with the animal before anaesthetizing.

Jaw traps (JT) : the model used was officially confirmed by the French Environment Agency and is actually used by French wildlife agencies and hunting federations to control marmot populations. The edges of the jaws were fitted by rubber. The trap measured 18 cm in diameter and was fastened to a peg by a 50 cm steel chain. The peg was placed in such a way as to prevent the marmot entering deeply into the gallery but enough to stay quiet in the dark. When extracted from the burrow the animal was grasped by the nape of the neck to remove the trap, and confined in a linen bag.

Snares (ST): as the pattern used by Italian wildlife agencies our snares consisted of flexible steel wire (bicycle brake wire) fitted with a safety locking mechanism stopping the closure to 8 cm in diameter (25 cm before the ending loop). The loose end of the wire was strongly fixed to a peg. The loop of the snare was placed inside the mouth of the burrow and held in position by small twigs. The length of the wire must be adjusted to prevent snared marmot from entering too deeply into the burrow. When caught, animals were handled in the same way as for jaw traps.

Nine trapping areas were defined in term of 3 main kinds of environments (sunny slope, shady slope, bottom of valley), each one divided in 3 sectors distant of 150 m from the next one.

Each week, 6 out of the 9 sectors were tested according to a rotating schedule in order to reduce the effects of weather and season and to set in evidence effects related to the kind of trap and/or to the kind of natural environment. The traps were installed on the first day of the session. They were set early in the morning of the second day until released at dusk. The same for days 3 and 4 at dusk of which traps were removed until next week.

Five trapping sessions occurred in 1992 on weeks : 22-24/April ; 28-30/April ; 07-09/May ; 12-14/May ; 18-21/May. Independently of these sessions, some snares were tested.

In 1993, 6 trapping sessions occurred on weeks : 03-05/May ; 12-14/May ; 17-19/May ; 27-29/May ; 04-06/June ; 10-12/June.

Traps were checked every hour in order to remove caught animals as soon as possible. The moment of each capture was recorded.

Trapped animals were anaesthetised when removed from the trap by an intra-muscular injection of Zoletil 100 (Reading Laboratory; Tilétamine base 250 mg + Zolazépam base 250 mg + distilled water 5 ml; 15-20 mg/kg). Once tranquilized they were sexed (Zelenka, 1965) and measured. Age classes were determined by combination of body weight and head and body length (Juan 1993) allowing to distinguish 3 classes : yearlings (born previous year), 2-years old, adults. As juveniles emerged only at the end of June they do not appear in that experiment.

We took into account the different injuries and losses that could be related to trapping activity. Out of unhurt animals (no observed mark which could be related to trapping), we have distinguished 4 levels of increasing injuries:

• Light traumatisms: small wounds which did not handicap the animal when released; pulled out claws; gashes on palms or muzzle; scratches on the nose.
  
• Hematomae and muscular bruises. Hematomae reached often the size of a hen egg on animals weighing 1.2 kg.
  
• Heavy traumatisms. The observed bone fractures concerned radius and cubitus as marmots were trapped by one foreleg.
  
• Death.

Marmots were released at the trapping point when totally aroused (around one hour later).

During this experiment, a total of 102 captures were realised in 979 trap-days.

Table 1. Efficiency of the different traps

1DT: one-door live trap; 2-DT: two-door live trap; JT: jaw trap; ST: snare trap. Efficiency is the number of marmots captured per trap and per day of trapping (11 marmots trapped per 135 trap/day).

No differences appeared between the efficiency of capture (number of marmots captured per trap and per day of trapping) of one-door and two-doors live traps whatever the year. However the efficiency of this two kind of traps increased in 1993 (comparing the same number of sessions of captures : 1992, 21/255; 1993, 38/255; Chi square test: c2 = 4.9, df = 1, p = .027).

In 1992 jaw traps were significantly more efficient than the two types of live traps (c2 = 7.1, df = 2, p = .013).

Live traps were more efficient than snares in 1993 (c2 = 12.9, df = 2, p = .0026). Snares showed important changes from one year (.26) to the other (.03), and the weakness of the frequencies observed in 1993 made difficult any further comparison. Snares were excluded of the following analysis.

The proportion of marmots trapped during the four periods of 3 hours and the two periods of six hours of the day was computed for each kind of traps.

The proportion of caught animals during the two six periods of the day was the same whatsoever the alternative trap (Coefficient of contingency: C = .081, df = 2, c2 = .57, p = .75). Nevertheless the maximum of trapped animals each day occurred for the three kind of traps significantly during the first three hours of the day (c2 = 39.9, df = 3, p = .0001).

Figure 1. Day-time trapping success (1992 + 1993)

Trapping frequency in each three hour period of daylight.

Each session of trapping lasted three days. Did the effect of novelty affect in some way the success of trapping of the three alternative traps? Even if there was an increase trend from the first day to the third, there was no significant relation between traps and the day of the session (Coefficient of contingency C = .192, df = 4, Chi square = 3.37, p = .49). The success of trapping of the different kind of traps did not differ from one day to another (df = 2; 1DLT: c2 = 3.2, p = .21; 2DT: c2 = 3.8, p = .15; JT: c2 = .75, p = .11).

Figure 2. Frequency day trapping success during sessions (1992 + 1993)

Trapping frequency in each three day of a trapping session.

The trapping periods were not exactly the same during the two years (table 2). A Chi-square test on the frequency of captures of each week showed no difference between weeks.

Table 2. Weekly captures

As traps were placed on three kinds of environment it is possible to compare capture frequency according to the exposition of the slope.

No interrelation appeared between the three alternative traps and the exposure of the slope (Contingency coefficient C=.24, c2=5.31, df=4, p=.26). The captures are significantly higher on sunny slopes than in shady slopes or valley bottom (c2=6.69, df=2, p=.036).

Figure 3. Effect of the exposition of the slope (1992 + 1993)

Trapping frequency in each type of slope exposition. Sun: Sunny side

As all animals were not sexed, in this case the total number observed is lower than the total captured.

The ratio males/males + females was considered as sex-ratio. On the whole captures the ratio was equilibrated (.51). Important variations were observed for one-door live trap (.35) and for jaw traps (.66) but no significant difference between the observed frequencies of males and females occured (Binomial test, P=Q=1/2, p>.05).

Table 3. Sex-ratio (1992 + 1993)

M: males; F: females ; Sex-ratio (S-R): Number of Males/Total.

As some uncertainty existed in age determination, only 78 captures are analyzed. Adults represented 50% of the total, 2-year old 33% and 1-year old 18 %. Snares trapped mostly adults (c2=10.3, p=.0013). On the contrary, one-door, two-door live traps and jaw traps captured the same proportion of animals of each class of age (Contingency coefficient C=.22, df=4, c2=4.26, p=.37). But the proportion of one-year old animals were smaller than that of two-year-old and adult ones.

Figure 4. Age classes (1992 + 1993)

Trapping frequency in each age class.

In a bulk, 70 % of the marmots were injured by jaw traps, less than 30 % by live traps and 22 % by snares.

Jaw traps produced severe injuries in 57 % of the captured animals including 4 % of death (one marmot). The reasons of the death of this animal were not clearly determined, but could be related to heat and/or stress. Among the fractures we noticed an open fracture.

Other traps did not produced heavy traumatisms to marmots.

Figure 6. Injuries and pattern of traps

None of the marmots captured by jaw traps were captured again by this trap. In 1992 six marmots were captured twice by live traps corresponding to 40 % (6 over 15 different trapped animals) of the animals captured this year by this kind of trap. In 1993, 10 individuals were re-captured (7 of them twice and 3 of them three times) corresponding to 38 % of the marmots captured by live traps (10 over 26) (intervals : 2 to 26 days). In most cases they were captured by live traps. In two cases they were also captured by a snare: one at the first capture, the other at the third capture.

The physical difficulties encountered in the use of the different pattern of traps and the cost of these traps are also criteria invoked by wildlife managers.

Weight: snares are small devices weighing 100-200 g each. Jaw traps weigh (1 kg) and 1-door and 2-doors live traps respectively 3.5 and 5 kg.

Bulkiness: the most reduced bulkiness is for snares, then jaw traps and at least one-and 2-doors live traps. A single trapper will commonly transport a great number of snares or around 15 jaw traps. But it will be difficult to carry more than 4 live traps at the same time particularly in steep environments.

The choice of a pattern of trap could be influenced by purchase cost of the different devices. Following prices are only indicative as subject to variations from one country to another. In France the price of a snare varies from 5 to 10 FF (1 to 2 US$). Jaw traps cost : 120 FF (20 $). One-door live trap is around 800 FF (160 $) and 2-doors live traps is 1000 FF (200 $). Live traps cost could be reduced by self manufacturing them. Out of cost of labour their price varies then from 50 to 100 FF (10 to 20 US$).

The cost of maintenance has also to be considered. Snares are more brittle than the other traps and will have to be changed more often. Jaw traps and live traps are long lasting devices and keeping them in good order is easy and can be done by the trappers.

Trap handling acts secondarily on the cost of a trapping operation as it determines the number of trappers or the time spent by trapper to set the traps in fields. It limits trapping in areas near tracks.

Significant differences were set in evidence in the efficiency of the different types of traps for capturing Alpine marmots. In table 4 we tried to highlight some of the positive and negative points set in evidence in our experiment.

Table 4. Comparison of the different pattern of traps

Framed cells: Positive points. Shades cells: Negative points.