Territory characteristics and coexistence with heterospecifics in the Dartford warbler Sylvia undata across
Yüklə 0,64 Mb.
|
* TO * * LJ * *  Sylvia undata
Sylvia melanocephala Sylvia cantillans * Overlap S.undata/S. undata Overlap S.undata/S. melanocephala Overlap S.undata/S. cantillans
Meters
 Table 3 Vegetation structure and territory characteristics of the Dartford warbler among plot-years
Plot-year Cv H′ Year Pairs/  10 ha
Locations Mean size SD size Max size Min size Roundness N LJ87 1.33 1 0.0 – – – – – – – LJ88 1.52 2 2.8 – – – – – – – LJ89 1.61 3 11.3 15.6 3,734 1,286 5,916 1,267 0.867 17 LJ90 1.62 4 16.4 15.8 2,597 1,162 4,689 1,142 0.841 27 LJ91 1.61 5 11.1 15.5 3,049 1,248 5,956 1,132 0.861 17 LJ92 1.58 6 9.5 18.7 5,348 1,981 9,495 3,056 0.882 15 LJ93 1.66 7 8.3 18.8 4,930 2,250 8,654 1,544 0.822 13 LJ94 1.62 8 7.3 17.8 5,530 1,789 8,857 3,256 0.858 10 TO91 1.36 15 10.0 25.9 5,418 1,563 8,292 3,237 0.884 9 TO92 1.41 16 9.7 18.8 4,716 1,468 6,165 1,738 0.854 9 TO93 1.40 1 5.5 44.4 9,255 5,100 15,033 3,760 0.857 5 TO94 1.31 2 7.6 26.6 7,507 3,696 16,140 4,243 0.834 8 TO95 1.28 3 8.8 31.9 8,557 2,933 13,638 5,531 0.864 8 TO96 1.29 4 11.5 17.7 4,615 1,986 7,529 1,379 0.834 10 TO98 1.34 6 17.3 23.0 2,611 1,486 6,063 574 0.813 16 TM97 1.50 38 7.1 38.5 10,071 3,527 17,121 6,267 0.903 8 TM98 1.50 39 6.9 16.9 5,532 2,466 9,531 2,369 0.849 9 TM99 1.50 40 6.5 9.9 5,061 1,963 6,965 2,214 0.836 7 TM02 1.19 1 0.8 10.0 4,075 – – – 0.907 1 TM03 1.20 2 2.4 10.3 7,905 1,907 10,085 6,543 0.847 3 TM04 1.19 3 2.0 9.0 4,553 1,503 5,616 3,490 0.869 2 TM05 1.16 4 3.3 9.0 5,063 1,410 6,353 3,557 0.823 3
Cv H′ structural diversity, year years after fire, pairs/10 ha breeding pairs of Dartford warbler/10 ha, locations mean number of bird locations per territory, size territory size (minimum convex polygon in m2 ), roundness perimeter circle/perimeter territory having the same area, N number of territories measured, – non-measurable. selection of variables. We found a small plot effect in the selected model for territory size. As was expected, the number of locations had a strong influence on the Dartford warbler’s territory size (Fig. 2). Once this effect had been accounted for in the models, the GLMM using density variables showed a clear reduction in territory size as conspecific density increased and distance to nearest neighbor decreased (Table 4). This relationship is apparent (Fig. 3a) when plotting the relative territory size (the residuals of regression between of the number of locations and territory size, both log10 transformed) against conspe-      4.3
  3.9
  3.5
   3.1
2.7
cific density. On the other hand, there was no evidence of a direct effect of heterospecific density on territory size, since it did not enter into the model once conspecific density had been considered. Cover variables did not enter into the habitat model although the effect of foliage cover at 0.25 to 0.5 m was nearly significant (P= 0.07), showing a negative relation- ship with the relative territory size (Fig. 3b). Accordingly, the overall model developed using all explanatory variables was equivalent to the density model.
0.85 (±0.07 SD), ranging from 0.56 to 0.97 (n = 197). Both forward and backward stepwise models selected territory size, in addition to the number of locations and plot, as the main explicative variable for territory shape (Table 4). Large territories were circle-shaped, while small ones showed a greater shape variation and often diverged widely from circularity (Fig. 4). In addition, the backward model 0.4 0.8 1.2 1.6 2 showed a significant positive effect of the conspecific Log10 number of locations density and a significant negative effect of the heterospe-
Fig. 2 Relationship between Dartford warbler territory size and the number of locations used to delineate territories cific density on territory roundness. Dartford warbler territories were more circular at high density of conspecifics
Table 4 Results from the GLMM of Dartford warbler territory size and roundness
df b SE F P Partial η2
For size, different models were run using bird density variables and habitat variables alternatively. Plot was included as a random factor. and at low density of heterospecifics, although this effect was modest. Intraspecific territory overlap was low (Fig. 5a), averag- ing 3.0% (±7.6% SD) of the total territory area, and ranging 0–49.4% (n = 197). Dartford warbler density did not have any significant effect on overlap (Table 5) and therefore intraspecific overlap can be considered density-indepen- dent. A significant, though rather weak, effect of the plot appeared in this model. No habitat variable had a significant effect on intraspecific overlap (Table 5). The mean interspecific overlap between the Dartford and the other three warblers (Sylvia and Hippolais species) was over a degree of magnitude larger than intraspecific overlap (mean 50.6 ± 44.0% SD, range 0–179.1%, n = 197, Fig. 5b). The plot had a significant effect in the habitat model. Both heterospecific bird density and structural diversity had significant and positive relationships with interspecific territory overlap, and the magnitude of both effects was similar. The overall model maintained both explanatory variables (Table 5), suggesting that interspecific overlap is a density-dependent phenomenon that can also be enhanced by certain habitat characteristics. Discussion
Fig. 3 Relationships between Dartford warbler relative territo- ry size (the residuals of regres- sion between the number of locations and territory size, both log10 transformed) and: a Dart- ford warbler density; and b relative foliage cover of the 0.25–0.5 m vegetation layer, the best although non-significant (p = 0.07) habitat variable
1
      -1 
   -2
-3 0 5 10 15 20 20 40 60 80 100 S. undata density (territories/10ha) 0.25-0.5m cover (%)  1
   0.9
0.8
0.7
0.6
0.5
  2.7 3.1 3.5 3.9 4.3
The measured territory sizes for the Dartford warbler lie within the range published for this species (0.14–2.5 ha) although the variety of data and methods used in territory drawing precludes direct comparisons, even when mini- mum convex polygons are used. The strong effect of the number of locations used to delineate polygons on territory size (Table 4) is an additional reason to prevent inter-study comparisons. The underestimation of territory size arising from the mapping method may be extreme at low population densities. Below 3 b.p./10 ha of the Dartford Log10 terr itory size (m2) warbler, the meager singing activity of territory owners at
Fig. 4 Relationships between Dartford warbler territory roundness and size
gested by a certain plot effect on our models. This effect may be due, after excluding very similar climatic con- ditions, to local differences in food availability or in plant community derived from different geology and topography (cf. Table 1). precluded the analysis of those territories (Table 3). This effect points out that the mapping method is constrained by the density-dependence of territorial behavior, a limitation that has been seldom analyzed but is paramount when studying skulking birds. Moreover, territory mapping of unmarked individuals probably underestimates intraspecific overlaps, because it is assumed that a unique territory boundary falls between simultaneous records of conspecific Fig. 5 Relationships of Dart- ford warbler territory overlap with bird density and habitat variables. Bird density in the intraspecific overlap scattergram (a) is Dartford warbler density, while the interspecific one (b) is overall density of shrub- dwelling warblers, S. undata excluded
180
120  90
    a Intraspecific b Interspecific
|
