Ardea
Official journal of the Netherlands Ornithologists' Union

login


[close window] [previous abstract] [next abstract]

Perdeck A.C. & Speek G. (1984) A radar study on the influence of expected ground speed, cloudiness and temperature on diurnal migration intensity. ARDEA 72 (2): 189-198
In The Netherlands there have been extensive field observations of the autumn diurnal migration. As a follow-up diurnal broad front migration was studied near Arnhem with a 3.18 cm pencil-beam radar (range 2.5 km) during 8 springs (134 days) and 7 autumns (140 days). This paper deals with the influence of the season on the relationship between migration and the weather. Migration intensity was assessed by counting the number of echoes per time unit per day from pictures on the radar screen. The wind factor was taken as the 'expected ground speed' (Alerstam 1976). The direction and speed of a bird (flock), can be calculated from consecutive echoes (afterglows on the screen). When there was no wind the mean speed was 16 m/sec and the mean direction was 55¦ in spring and 22SO in autumn. These values give an estimate of the mean air speed and the mean preferred direction. The wind vector (W) was assessed with balloons followed on the radar screen. The expected ground speed (T) in the preferred direction can then be calculated (Fig. 1), after it has been corrected for the (according to the unknown species) varying air speed. With known T and W, H can be found under different values of W. When H is related to the wind component in the track direction (y) it appears to vary with the find speed, confirming the findings of Bloch et al. (1981) that the air speed decreases with tail winds, but increases with head winds (see under Fig. 1). The expected ground speed was corrected for this relationship. Two other weather factors were taken into consideration, viz. cloud-cover and temperature. The difference between spring and autumn with regard to the relationship between migration intensity and the three weather factors was studied in a multivariate non-parametric way by means of a loglinear analysis of multidimensional contingency tables. The results are as follows: 1. The wind (expected ground speed) has a positive influence on migration intensity, both in spring and autumn. In both seasons birds use the wind to maximize their efforts. This is a clear affirmation of the ideas that originated during the field observations. 2. During both seasons, cloud cover has a negative influence on migration intensity. This confirms the results of studies on orientation that revealed that the visibility of the sun is a factor in the orientation of birds. 3. In spring high temperatures stimulate migration. confirming findings from other studies. Temperature in the area of observation may serve as an indication of temperature in the breeding area. Birds are thus prevented from starting to breed under adverse conditions. 4. In autumn temperature has no influence on migration intensity. This is contrary to the general idea that falling temperatures stimulate autumn migration. Low temperatures and tail winds (NE winds, high expected ground speed) often occur together during autumn (anticyclonic weather type). The statistical technique that was used made it possible to separate such independent factors, and points to the expected ground speed as the causal factor. This result is confirmed by observations in the Baltic (Zalakevicius 1982) where the influence of temperature showed a similar difference between autumn and spring. It may be concluded that the radar observations generally confirm the results of the Dutch field observations, and that these are now extended to the situation during spring.


[close window] [previous abstract] [next abstract]