Microtus arvalis

Balčiauskas, Linas, Received, Laima Balčiauskienė & Online, Published, 2014, Selective predation on common voles by Tawny Owls and Long-eared Owls in winter and spring, Turkish Journal of Zoology 38 (1), pp. 242-249 : 243-246

3.1. Regressions based on trapped M. arvalis

We found that in trapped voles all measured cranial characters significantly correlated with the body mass, yielding regression equations with various predictability ( Table 2).

In general, the average of the measured body mass of trapped M. arvalis was 17.73 ± 0.49 g (10.5–32.0 g in 2005 and 2007–2009; 7.8–33.5 g in 2006), while the average of the calculated body mass was 17.14 ± 0.14 (14.6–22.1) g. The difference is not statistically significant, with an error of 3.3%; thus, regressions accurately predict body mass.

We calculated differences between measured and calculated body mass and expressed it as the ratio to the average body mass of the trapped individuals. Comparing single characters, just one regression (based on X 17) showed a statistically significant difference between the measured body mass and that obtained from regression. The calculated body mass was underestimated in most regressions by 0.9%–6.0%, with the exception of the regression based on X 2, which yielded a 0.7% overestimation ( Table 3).

3.2. Body mass of trapped and predated M. arvalis Factorial ANOVA showed that the effect of the sampling type (trapping, preying of S. aluco , and preying of A. otus ) on the body mass of M. arvalis is significant (F 1, 1321 = 186.26, P <0.0001), while effect of the season is not. Interaction of both factors is significant (F 1, 1321 = 7.36, P <0.01); thus, the effect of sampling type in winter and spring is not the same. Furthermore, we analysed these differences in detail.

From trapped M. arvalis voles, it was found that the average body mass of juveniles was 14.53 ± 0.07 g (7.8– 22.3 g, n = 378), of subadults was 18.07 ± 0.27 g (13.0– 25.1 g, n = 81). and of adults was 22.5 ± 0.37 g (13.3–33.5 g, n = 100), with an average irrespective of age equal to 16.42 ± 0.16 (7.8–33.5) g. Clearly, the distribution is biased towards individuals with small body mass (i.e. young individuals), prevalent in the population in most of the nonvegetative period. The body mass distribution of trapped M. arvalis individuals is shown in the Figure 1 (recalculated from Balčiauskienė et al., 2009).

The average body mass of M. arvalis preyed upon both by S. aluco and A. otus was 21.72 ± 0.12 (14.3–29.3) g when estimated from crania (n = 366) and 21.45 ± 0.12 (9.2–30.4) g when estimated from mandibles (n = 497); the difference is not significant. Having in mind better preservation of mandible characters and larger number of recovered mandibles ( Table 2), absence of body mass difference made it possible to rely on results calculated from mandibles. Through comparison with the trapped individuals in the nonvegetative period, such an average body mass of predated individuals indicates that adult voles are being predated ( Figure 1 ).

Comparison of the average body masses of the trapped and predated M. arvalis shows that predated individuals were significantly heavier by over 30% of body mass than trapped individuals. Predated M. arvalis were 5.03 g heavier; the difference is highly significant (t = 25.87, df =

140

individuals 120 100

80

of

Number 40 60

20

1064, P <0.001). The underestimation of the body mass by most regressions (see Table 3) makes this difference even greater.

3.3. Prey preferences

Owl-predated voles were not only the heaviest individuals in the population ( Figure 1 ); body mass also reflects different exploitations of the population of M. arvalis . In terms of age structure, juveniles (up to 16 g) were highly under-predated; they accounted for 71.8% of the population but only 3.4% in the prey items (the difference highly significant, χ 2 1 = 516.9, P <0.001). The proportion of subadult voles (body mass 17–19 g) in the tested population was 12.3%, while in the prey it was 25.2% (χ 2 1 = 29.5, P <0.001). Adult voles were highly over-predated; while they accounted only for 16.0% of the sampled population in the nonvegetative period, they made up 71.4% of the prey (χ 2 1 = 335.73, P <0.001).

We found that the differences between the average body masses of the prey in the winter-spring diet of A. otus , the winter diet of S. aluco , and the spring diet of S. aluco were all significant (F 2, 494 = 6.42, P <0.001). Within each of these groups, the difference of predated M. arvalis body mass calculated from crania and mandibles was not significant (Wilks’ lambda = 0.97, F 4, 724 = 2.33, NS).

In winter and spring, A. otus preferred larger individuals of M. arvalis (average body mass 21.56 ± 0.11 g) than S. aluco did in winter (19.59 ± 0.66 g); the difference is significant (t = 4.51, df = 477, P <0.001). However, the largest voles were preyed upon by S. aluco in spring (average body mass 22.39 ± 0.70 g); i.e. the S. aluco prey was bigger in spring than in winter (t = 2.63, P = 0.01). Within this spring group, breeding S. aluco preyed upon larger voles than nonbreeding ones (21.80 ± 1.08 g).

The body mass distribution of M. arvali s preyed upon in winter by S. aluco and winter-spring by A. otus was significantly different ( Figures 2A and 2B ). In S. aluco prey, the proportion of young voles was significantly higher than in the prey of A. otus : respectively, 33.3% versus 5.2% (χ 2 1 = 31.47, P <0.001). By contrast, the proportion of subadult and adult voles with a body mass in the range between 19 and 23 g was significantly higher in the prey of A. otus , respectively, at 72.9% versus 55.6% (χ 2 1 = 5.95, P = 0.014).

The distribution of the body mass of M. arvalis preyed upon by S. aluc o in spring ( Figure 2C ) was biased towards the heaviest individuals in the population, i.e. adults. Individuals with a body mass of over 24 g accounted for 44.4% of the predated voles. This is significantly more than the proportion of the heaviest individuals in the winter-spring prey of A. otus (12.9%, χ 2 1 = 5.64, P <0.02) and in the winter prey of S. aluco itself (16.0%, χ 2 1 = 7.71, P <0.01). In addition, the proportion of heaviest individuals in the spring prey of nonbreeding S. aluco (4 out of 9 individuals) was also higher than that in the winter-spring prey of A. otus (χ 2 1 = 5.15, P = 0.02) and in the winter prey of S. aluco (χ 2 1 = 5.47, P <0.02).

Figures

Figure 1. Distribution of the body mass of Microtus arvalis individuals, trapped in October–April and preyed upon by both Strix aluco and Asio otus in winter and spring, as estimated from mandibles.

Figure 2. Distribution of the body mass of Microtus arvalis individuals, preyed upon by Strix aluco and Asio otus in winter and spring 2004–2005, as estimated from mandibles: A) in the winter-spring diet of Asio otus; B) in the winter diet of Strix aluco; C) in the spring diet of Strix aluco.