The Auk 123(2):537–547, 2006 © The American Ornithologists’ Union, 2006. Printed in USA.

DO PARTIAL AND REGULAR MIGRANTS DIFFER IN THEIR RESPONSES TO WEATHER? A L. K. N,1 T A, J-Å N Department of Animal Ecology, Lund University, Ecology Building, 223 62 Lund, Sweden

A.—Partial migration is o en considered a transitory stage between migration and residency, and whether partial migrants take weather conditions into account during migration is largely unknown. To assess whether partial migrants diff er from regular migrants in their responses to weather, we compared the migratory intensity of a partial migrant, the Blue Tit (Cyanistes caeruleus), with more regular migrants in relation to weather at a migratory passage site in southern Sweden (Falsterbo) during the years 1993–2002. The regular migrants in the study were Linnet (Carduelis cannabina), Common Chaffi nch (Fringilla coelebs), Brambling (F. montifringilla), and European Robin (Erithacus rubecula). The Blue Tit diff ered from the regular migrants mainly in showing a striking negative correlation between migratory activity and cloud cover. Also, weather had the highest explanatory power for migratory intensity in the Blue Tit. This suggests that the Blue Tit is more sensitive to weather conditions on migration than the regular migrants and that it preferably awaits days with wholly or partly clear skies before migrating past Falsterbo. As a consequence, Blue Tits usually restrict their migratory fl ights to the safest occasions, with relatively calm weather, good visibility, and all orientation cues (solar as well as magnetic) available. Received 29 April 2004, accepted 20 August 2005.

Key words: Blue Tit, Cyanistes caeruleus, evolution of migration, partial migration, timing of migration.

¿Difi eren los Migrantes Parciales y Regulares en sus Respuestas al Clima?

Rz.—Comúnmente se considera que la migración parcial es una etapa transitoria entre la migración y la residencia, y se desconoce en gran medida si los migrantes parciales tienen en cuenta las condiciones climáticas durante la migración. Para evaluar si los migrantes parciales difi eren de los migrantes regulares en sus respuestas al clima, comparamos la intensidad migratoria de un migrante parcial, Cyanistes caeruleus, con la de migrantes más regulares con relación al clima de un sitio de paso migratorio en el sur de Suecia (Falsterbo) durante los años 1993–2002. Los migrantes regulares de este estudio fueron Carduelis cannabina, Fringilla coelebs, F. montifringilla y Erithacus rubecula. C. cannabina se diferenció de los migrantes regulares principalmente por mostrar una correlación marcadamente negativa entre la actividad migratoria y la cobertura de nubes. Además, el clima tuvo el máximo poder explicativo para la intensidad migratoria en C. cannabina. Esto sugiere que C. cannabina es más sensible a las condiciones del clima durante la migración que los migrantes regulares y que preferentemente aguarda los días con cielo total o parcialmente despejado antes de migrar a través de Falsterbo. Como consecuencia, C. cannabina usualmente se limita a realizar sus vuelos migratorios bajo las condiciones más seguras, con tiempo relativamente calmo, buena visibilidad y con todas las señales de orientación disponibles (tanto solares como magnéticas).

1E-mail: [email protected] 537 538 N, A, N [Auk, Vol. 123

F {z~ perspective, it may be migrate in winds from the southwest (i.e. head somewhat tricky for a sedentary bird species to winds)—which, they argued, was a ributable become migratory. Success as a migrant requires to poor fl ight ability and, therefore, an adapta- expression of several genetic components (Dingle tion to avoid wind dri over the sea. 1996). Because migration is energetically costly, Our aim here is to investigate the behav- migrants must store fuel before migration. They ior of Blue Tits in comparison with regular must be able to determine timing, distance, and migrants migrating during the same period, direction of migration (Helbig 1996, Berthold with respect to the relationship between 2001), as well as to orient themselves and navi- weather and migratory intensity. Our main gate. Furthermore, migrants must be able to hypothesis is that partial and regular migrants respond to atmospheric conditions by timing diff er in their responses to weather, which can their departures so as to avoid traveling in unfa- be expected if partial migration represents a vorable or dangerous weather (Alerstam 1978; transitional evolutionary stage between resi- Richardson 1978, 1990; Kerlinger and Moore dency and migration. Such diff erences might 1989) and to reduce migration costs by taking arise because adaptations to migration have not advantage of favorable weather. Migrants are been perfected in partial migrants, which are o en observed fl ying in tailwinds (Gauthreaux therefore constrained in their migratory ability. and Able 1970), which reduces the energy costs Furthermore, partial migrants have to deal with and duration of their trip and, thus, their expo- more complex selection pressures, because they sure to predators and other dangers encountered also need adaptations for residency. on the way (Alerstam 1978; Richardson 1978, If partial and regular migrants diff er in rela- 1990). Precipitation, another important weather tion to weather, partial migrants may be con- factor, is commonly associated with low migra- strained in the ability to travel under variable tory intensity (Alerstam 1978, Åkesson et al. and diffi cult conditions, or in the ability to select 2001, Dänhardt and Lindström 2001). It has been the most favorable weather for migration. Hence, suggested that precipitation increases the costs we will fi rst test our hypothesis about the exis- of migration by chilling migrants and increas- tence of diff erences in weather response between ing their mass (Alerstam 1978, Richardson 1978). partial and regular migrants. If such diff erences Poor visibility is also known to aff ect migratory are found, we will evaluate whether they indi- intensity negatively, though some species are cate that Blue Tits diff er from regular migrants observed migrating beneath, above, or even in mainly in their ability to choose the most favor- clouds (Richardson 1990). able conditions for their migratory fl ights or in Partial migration, in which only a fraction of their ability to continue migration under diffi cult a population—o en subdominant individuals conditions (e.g. under an overcast sky, lacking such as juveniles and females—is migratory (e.g. solar orientation cues; in strong winds that cause in the Blue Tit [Cyanistes caeruleus; Smith and problems of wind dri and compensation; or Nilsson 1987]), is o en considered a transitional in poor visibility, which reduces the potential stage between residency and regular migration. use of landmarks; etc.). The fi rst case may be Do partial migrants—in which the individual’s inferred if partial migrants show smaller varia- readiness to depart on migration is conditional— tion than regular migrants in migratory intensity respond to weather in the same way as regular between days with diff erent weather conditions migrants, which are adapted to perform annual and if weather accounts for less of the variation migratory journeys throughout their lifetimes? in migratory intensity among partial than among The Blue Tit is a common species, occurring regular migrants. The second case may apply year-round in forested habitats over large parts if partial migrants tend to travel under specifi c of Europe. However, in the northern part of its weather conditions and if weather explains more range (i.e. Scandinavia), the Blue Tit is a partial of the daily variation in daily migratory intensity migrant and frequently appears both in ringing than in regular migrants. statistics and visual observations at Falsterbo, on the southwestern tip of Sweden (Heldbjerg and M Karlsson 1997). In terms of its weather response during autumn migration, Lindskog and Roos The study was done using the visual obser- (1980) found that the Blue Tit preferred to vation counts and daily ringing records of April 2006] Weather Response in Partial Migrants 539 migrating birds at Falsterbo, in southwestern Wind and weather conditions are recorded Sweden, for 10 years, 1993–2002. Selected study by the Swedish Meterological and Hydrological species from the visual observation material, Institute (SMHI) at the weather station in apart from the Blue Tit, were Linnet (Carduelis Falsterbo lighthouse garden. For wind direction cannabina), Common Chaffi nch (Fringilla coe- and speed, barometric pressure, temperature, lebs), and Brambling (F. montifringilla); Common visibility, cloud cover, and humidity, observa- Chaffi nches and Bramblings regularly travel tions are recorded every 3 h. Cloud cover was in mixed fl ocks and cannot be counted sepa- recorded on a scale of 0–8 units, where 0 is rately). From the ringing material, we selected clear and 8 overcast. For the present study, the European Robin (Erithacus rubecula) and the weather recordings at 1000 hours were used in combined captures of Common Chaffi nch and all analyses. Precipitation is recorded during Brambling as representing regular migrants. 12-h intervals, from 0600 to 1800 hours and In southerly parts of their breeding ranges, from 1800 to 0600 hours. The diurnal record the European Robin and Common Chaffi nch of precipitation was used for the present study. are partial migrants, though they are generally Because of the problem of treating circular and not considered as such in Sweden. The regu- continuous data in the same analysis, wind lar migrants in Falsterbo are mainly obligate direction, measured in degrees, was trans- migratory populations from Scandinavia and formed into a category variable refl ecting the more northeasterly areas. The Brambling, for four quadrants: northeast, southeast, south- instance, breeds exclusively in the mountain west, and northwest winds. ranges in northernmost Scandinavia. Therefore, We treated and analyzed visual observations we do not think that the improbable and mini- and ringing separately, because they represent mal occurrence of partially migratory popula- diff erent aspects of migratory activity. Birds cap- tions among our regular migrants will aff ect tured for ringing have interrupted their migra- the aims of the present study. These regular tion, in contrast to birds recorded by visual migrants are all common, short- to medium- observations, which are seen migrating past the distance, small passerine migrants, occurring study area. For the two species, where compari- at Falsterbo during approximately the same sons of the two recording methods are possible, period as the Blue Tit and thus exposed to the the correlation was not consistent (Blue Tit: r = same variation in daily weather conditions (for 0.72, P < 0.001; Common Chaffi nch–Brambling: median dates, see Table 1). Like the Blue Tit, the r = –0.05, P = 0.5), which indicates that the two Linnet, Common Chaffi nch, and Brambling are data sets represent somewhat diff erent aspects diurnal migrants, whereas the European Robin of migration. is a nocturnal migrant. We are aware that neither the visual obser- Visual observations of migrating birds were vations nor the capture data from our study conducted daily from 1 August to 20 November site fully refl ect the true variation in migratory each year by one or two observers located at the intensity. Visual observations will be biased southwestern tip of the Falsterbo Peninsula. toward low-altitude fl ights along coastlines Observations began at dawn and lasted 8 h. (Alerstam 1978), whereas “captures” are birds Observations from 1993 to 2000 were carried that have temporarily interrupted their migra- out by Gunnar Roos, and those from 2001 to tory fl ights to land in the lighthouse garden. 2002 by Nils Kjellén. Still, these general biases are the same for the Migratory birds are mist ne ed daily from diff erent species, so that interspecifi c diff er- 21 July to 10 November in Falsterbo light- ences in weather relationships revealed by a house garden by Falsterbo Bird Observatory. comparative analysis are expected to refl ect Depending on weather and wind conditions, fundamental diff erences in departure activity in ≤21 mist nets are used. The nets are opened at relation to weather, in addition to possible dif- dawn and are closed, at the earliest, 6 h later, ferential weather eff ects on fl ight and landing or until <10 birds are captured during an hour. behavior. Between 75% and 80% of migrating Blue Tits Daily numbers of birds recorded in the visual are captured within the fi rst 6 h (Heldbjerg and observations (observation counts) and captured Karlsson 1997). Mist nets are emptied every half in the lighthouse garden (captures) were ana- hour, and more o en in damp weather. lyzed in relation to weather data using EXCEL 540 N, A, N [Auk, Vol. 123 b erent nch– 90 111 367 220 11.6 Common 44.1 b

7.5 683 821 European Chaffi 36.0 a

8.7 22.5 b

nch– 11.3 19.7 115,248 138 391,283 3,098 Common Chaffi counts Captures counts b

5.5 24.1 6,067 6,578 Linnet Brambling Blue Tit Robin Brambling 14,701 519,461 3,171 2,176 < 0.05) between species. < 0.05) between erences ( P erences a

19 30 36 39 49 65 4.8 275 10.9 Blue Tit 30,608 37,600 120,005 27,823 1,490,542 9,957 3,230 7 October 5 October 7 October 8 October 26 September 3 October 11 October 12 October 16 October 16 October 1 October 10 October cant diff cant 29 September 30 September 28 September 2 October 21 September 18 September deviation 5 8 9 9 11 16 11 9 deviation 5 8 9 superscript le ers indicate signifi superscript le of each species in the study based on annual data (1993–2002). Average migration period was tested with an ANOVA (Tukey HSD). D iff (Tukey ANOVA tested with an migration period was Average of each species in the study based on annual data (1993–2002). Maximum Minimum 29 25 10 7 13 6 Standard

T 1. Average annual number of migrating birds, median date, and duration of migration period for observed and captured birds, respectively, and captured birds, respectively, annual number of migrating birds, median date, and duration migration period for observed Average 1. T Observation deviation Standard Median date of migration Maximum Number of migrating birds Minimum Median Minimum Migration period (in days) Average Maximum Average Average Standard deviation Standard April 2006] Weather Response in Partial Migrants 541

2002 (Microso , Redmond, Washington) and combined in multivariate analyses, wind direc- SPSS, version 11.0 (SPSS, Chicago, Illinois). In tion and speed had a large eff ect on all species, the visual observations, one year of Blue Tit except for Common Chaffi nch–Brambling in the migration (1998) had to be excluded because ringing material, which are captured indepen- of negligible numbers of passing migrants. dently of all weather factors (Table 3). As for the Both univariate and multivariate general linear other weather variables, cloud cover seemed to models were used for assessing the absolute aff ect Blue Tit migration to a large extent. The and combined importance, respectively, of the migratory intensity of Blue Tits declined sharply weather variables. The dependent variable, with increasing cloud cover according to both daily migratory intensity, was expressed as pro- visual observations and captures, though the portions of the annual total, to correct for the decline was not as steep for captures as for visual annual variation in total numbers and to facili- observations (Fig. 2). The migratory intensity of tate comparisons between species. The annual European Robins also declined with increasing duration of the migratory period (number of cloud cover; however, the decline was more days) was defi ned as the 80% window of the gradual. Remaining species seem to migrate total migration in each season. Excluding the independently of this weather factor. In the data initial and fi nal 10% of the annual total of each set based on visual observations, visibility also species reduced the eff ect of extremely early explained part of the total variation in migra- and late migrants in our analyses. tory intensity in all species. In this interspecifi c comparison, Blue Tits showed the highest overall Rz dependence on weather, as indicated by the fact that weather explained the largest part of the The average annual number of recorded variation in migratory intensity of all species in birds and timing of migration of the selected both methods of recording migration (Table 3). species during the period 1993–2002 are pre- sented in Table 1. The species migrated during Dz approximately the same period, with median dates of migration occurring within the week In conclusion, we found Blue Tits to have of 3–8 October, with the exception of European a shorter migratory period than the regular Robins, which had a median date of 26 migrants in our study. Furthermore, migration September (Table 1). Blue Tit migration lasted of Blue Tits was more concentrated to peak for a signifi cantly shorter period than the other days, compared with that of regular migrants. species’ migrations, both for visual observations Blue Tits consequently had higher proportions (analysis of variance [ANOVA]: F = 7.26, df = 2 of days with li le or no migratory activity, and and 26, P = 0.003) and for captures (ANOVA: F = the days with intense migration were of greater 13.43, df = 2 and 27, P < 0.001). numerical importance for them than for regular Compared with the other species, Blue Tits migrants, most of which migrated on days of have a low proportion of days with moderate intermediate migratory intensity. We also found migration during the migration period. Instead, a striking negative correlation with cloud cover they have a high proportion of days with li le for Blue Tits alone, apart from correlations with or no migration as well as a higher proportion wind speed. Finally, weather explained a higher of days with intense migration than the regular proportion of variation in migratory intensity in migrants (Fig. 1A, B). Thus, a large proportion Blue Tits than in the other species. of Blue Tits migrate during specifi c days (Fig. Length of the migratory period.—The migratory 1C, D). By contrast, the largest cumulative num- period of Blue Tits at Falsterbo was shorter and ber of individuals of the other species tended to less variable than that of regular migrants. Li le be associated with days of intermediate migra- variation in the timing of Blue Tit migration tory activity. was also observed by Winkler (1974) at Col de The species varied in how weather variables Bretolet, and he concluded that the migration aff ected migratory intensity (Table 2). According pa ern resembled that of a regular migrant. to univariate analyses, Blue Tit migration The comparatively short and synchronized showed a signifi cant variation with wind speed, migratory period of Blue Tits may arise as a but not with wind direction. By contrast, when consequence of limited recruitment areas for 542 N, A, N [Auk, Vol. 123

Fl. 1. Proportion of days with different migratory intensities of each species (A and B) and the proportion of all migrants migrating during days with different migratory intensities (C and D). (A) and (C) refer to migratory intensity according to visual observations, and (B) and (D) to migra- tory intensity reflected by captures. Data for Blue Tit are shown by filled squares, for Common Chaffinch–Brambling by open triangles (dotted lines), for Linnet (only A and C) by open diamonds (broken lines), and for European Robin (only B and D) by open circles. the individuals of this species passing Falsterbo, to more distant regions in northern Scandinavia compared with possibly more wide-ranging and Finland, as demonstrated by the signifi cant recruitment areas for the regular-migrant spe- proportion of Bramblings (originating from cies. The origin of Blue Tits passing Falsterbo these northerly regions) among the migratory was analyzed by Heldbjerg and Karlsson Common Chaffi nches (see also Roos 1984). (1997), who concluded that “there should be An additional, and not mutually exclusive, li le doubt that most Blue Tits migrating at explanation for the short and synchronized Falsterbo originate from the southern third of migratory period of the Blue Tit may depend on Sweden.” Unfortunately, there are very few the species’ specifi c weather response, result- ringing recoveries to indicate the recruitment ing in migratory Blue Tits being inclined to areas of the other species (Roos 1984). However, leave on the fi rst opportunities with favorable these recruitment areas are likely to extend also weather within the migration window (see April 2006] Weather Response in Partial Migrants 543 a P nch– Common Chaffi

European < 0.010, signifi cance levels are given are given cance levels < 0.010, signifi P

Blue Tit Robin Brambling

a

P tested by two- species were response between erences in weather nch– Common counts Captures Chaffi

< 0.001. -values given for the species*weather interaction. For 0.05 < for the species*weather given -values P Blue Tit Linnet Brambling < 0.01, and *** P < 0.05, ** P cients are given except for wind direction, where one-way ANOVA was used to analyze diff erences in migratory used to analyze diff was ANOVA except for wind direction, where one-way cients are given species. erence between coeffi four groups of wind directions. Diff intensity between way ANOVA, with ANOVA, way as (*); * P Diff a Visibility 0.18 (*) 0.06 –0.03 0.090 0.10 0.09 –0.05 0.093 –0.05 species. Correlation and migratory intensities within between variables weather analyses between 2. Univariate T 0.09 0.10 0.090 –0.03 0.084 0.06 0.022 0.02 0.03 –0.15* –0.15* –0.02 direction Wind (*) –0.06 0.02 0.082 (*) speed Wind Pressure –0.13* ns 0.09 0.02 0.01 0.003 Temperature –0.11 (*) 0.27** 0.26** cover <0.001 Precipitation –0.18 –0.38** –0.46** 0.18 Visibility –0.07 0.16* –0.05 Cloud 0.00 <0.001 –0.23* –0.13* –0.06 0.029 Humidity –0.24* –0.07 0.20 (*) *** –0.11 (*) –0.21** *** –0.06 <0.001 0.009 –0.32** ns –0.35** 0.16* ns 0.02 0.22** <0.001 0.00 ** 0.012 ns 0.008

variable Weather Observation Observation 544 N, A, N [Auk, Vol. 123

T 3. Multivariate analyses of migratory intensity in relation to weather variables. Numbers show the proportion of the total variation explained by each weather variable (only statistically signifi cant contributions to explanatory power are included). Numbers do not add up exactly to the total explanation (r2) because of an imbalance in the ANOVA model (type III sum of squares was used to calculate the relative contribution of each variable to the full model).

Observation counts Captures Common Common Chaffi nch– European Chaffi nch– Weather variable Blue Tit Linnet Brambling Blue Tit Robin Brambling Wind direction 0.10 0.14 0.12 0.03 0.04 – Wind speed 0.15 0.04 0.08 0.09 0.11 – Pressure – 0.02 – – – – Temperature – 0.04 0.03 – – – Precipitation – – – – – – Visibility 0.05 0.04 0.02 – – – Cloud cover 0.13 – – 0.01 – – Humidity – – – 0.02 – – r2 0.447 0.275 0.250 0.220 0.163 0.000 r2 (with interactions) 0.447 0.347 0.293 0.247 0.190 0.013

below). Furthermore, because Blue Tits migrate the shortest distance of the species studied, they may not need as much time to prepare for migration as species migrating longer distances and can thus start migration on shorter notice. Degree of infl uence of weather on daily variation in migratory intensity.—The peak days of migra- tion are of great numerical importance in Blue Tits (Fig. 1). Whereas the regular migrants pass Falsterbo at a more constant rate over a longer period, Blue Tits leave during a few days with high migratory intensity (see also Lindskog and Roos 1980). We found that weather explained more of the variation in Blue Tit migratory intensity than in that of regular migrants, irrespective of recording method and associated biases (see above). The low explanatory power of weather in the case of regular migrants may refl ect a high level of adaptation to migration, assum- Fl. 2. Relationship between migratory inten- ing that regular migrants are able to migrate sity and cloud cover (mean migratory intensi- irrespective of most weather conditions. Blue ties per day and cloud cover category) for Blue Tits, on the other hand, seem to be more sensi- Tit (squares; filled for visual observations, open tive to weather conditions on their migration; for captures), European Robin (open circles, they await favorable weather before migrating broken and dotted lines), Common Chaffinch– past Falsterbo, which results in their migrating Brambling (triangles; filled for visual observa- during a few days with optimal conditions. This tion, open for captures, dotted lines), and Linnet higher sensitivity to weather indicates that Blue (filled diamonds, broken lines). Tits diff er from regular migrants in their ability to continue migration under diffi cult weather conditions and not in their ability to select the April 2006] Weather Response in Partial Migrants 545 most favorable conditions for their migratory mechanisms are present in these migrants is, fl ights (see above). Thus, partial migrants like to our knowledge, unknown (Wiltschko and the Blue Tit may not have the same ability as Wiltschko 1995). Another possible explanation regular migrants to continue migration in diffi - is that migratory movements are particularly cult weather. They are evidently more selective risky for Blue Tits and other partial migrants, of specifi c weather factors, such as cloud cover which are poorly adapted for long and sus- and wind direction and speed, which are impor- tained fl ights (and perhaps have small fuel tant predictors of Blue Tit migratory intensity. reserves; see Alerstam and Lindström 1990, However, the Blue Tit probably migrates the Kullberg et al. 1996). As a consequence, they shortest distance of the study species; perhaps usually restrict their migratory fl ights to the it can therefore aff ord to wait for favorable safest occasions, with relatively calm weather weather before departing on migratory fl ights. and good visibility and with all orientation cues Responses to diff erent weather factors.—Weather (solar as well as magnetic) available. cues used for migratory movements seemed In their responses to weather, Blue Tits also to diff er between our study species. All spe- resemble European Robins (capture data set in cies depended on wind direction (preferably Table 2), though the la er are exclusively noctur- migrating in headwinds; see Lindskog and Roos nal migrants and appear at Falsterbo lighthouse 1980) and speed, and to some extent also visibil- garden in the early morning, whereas Blue Tits ity (Table 3), which indeed are crucial for suc- generally appear before noon but well past sun- cessful migration (Alerstam 1978, Richardson rise (Heldbjerg and Karlsson 1997). Furthermore, 1978), but Blue Tits depended much more on European Robins are solitary migrants, whereas cloud cover than any of the other species (Fig. Blue Tits migrate in fl ocks. Despite these diff er- 2). This fi nding is supported by Lindskog and ences, both species are infl uenced by barometric Roos (1980), who also found a negative rela- pressure, precipitation, cloud cover, and wind tionship between migratory intensity of Blue speed in the univariate analyses. The resem- Tits and cloud cover. Alerstam (1978) analyzed blance may result from the method of sampling the daily migration counts of 15 species at (i.e. ringing) and its associated biases. However, Falsterbo in relation to weather. Two species it is still remarkable that the Blue Tit, in its were particularly strongly infl uenced by cloud weather response, seems to resemble a noctur- cover (increasing cloud cover being associated nal migrant more than the diurnally migrating with reduced migration), namely Hooded Crow Common Chaffi nch and Brambling. Although (Corvus cornix) and Eurasian Jackdaw (C. mon- the European Robin is well known to possess edula). As with Blue Tits, the variation in daily a magnetic compass (Wiltschko and Wiltschko migratory intensity of these two species could 1972), it also uses its celestial compass on migra- largely be explained by weather factors (r2 = tion during extended periods, between occa- 0.50 and 0.56 for Hooded Crow and Eurasian sions of calibration to geomagnetic directions Jackdaw, respectively; compare table 6 in (Wiltschko and Wiltschko 1975). Alerstam [1978]). It may be more than coinci- Evolution of weather responses.—In partial dental that both these species—like the Blue Tit, migrants, the evolution of migration and of which was not included in Alerstam (1978)—are responses to weather may diff er depending on distinctly partial migrants in Scandinavia, in the evolutionary pathway, which may have two contrast to the other species, which are mainly diff erent starting points. First, partial migra- or entirely regular migrants. tion may develop from sedentary populations Dependence on cloud cover in partial that expand into regions of stronger seasonal- migrants may be a specifi c weather reaction for ity, whereupon parts of the populations start to this sort of migratory behavior and an indica- migrate. Partial migrants among such typically tion that partial migrants have less ability to sedentary species include Blue Tits, Hooded continue migration under diffi cult weather Crows, and Eurasian Jackdaws. Furthermore, conditions, compared with regular migrants. the number of migratory Blue Tits has increased One explanation, albeit speculative, could be at Falsterbo during the past two decades that partial migrants primarily rely on the sun (Heldbjerg and Karlsson 1997). The selection compass for their orientation (Kramer 1957, pressures for the evolution of weather responses Emlen 1975). Whether magnetic orientation in such populations are entirely unknown, and 546 N, A, N [Auk, Vol. 123 it is largely unknown to what extent sedentary given from the Swedish Research Council (to birds are equipped with the necessary migratory T.A. and J-Å.N.). This is Report no. 225 from genetics. However, one interesting case is the Falsterbo Bird Observatory. House Finch (Carpodacus mexicanus), analyzed by Able and Belthoff (1998). House Finches from Lz C sedentary western populations were introduced to eastern North America, where a partially A, K. P., J. R. Bkk. 1998. Rapid migratory system rapidly evolved. Suppressed ‘evolution’ of migratory behaviour in the migratory traits, as in the House Finch, can lead introduced House Finch of eastern North to rapid evolution of migration, in contrast to America. Proceedings of the Royal Society evolution of migration by gene fl ow or muta- of London, Series B 265:2063–2071. tions, which is presumably a much slower pro- Å, S., G. W, L. K, S. cess (Pulido et al. 1996). E. 2001. Reed Warbler orientation: Finally, partial migration may also evolve Initiation of nocturnal migratory fl ights in in a regular migrant—for example, when relation to visibility of celestial cues at dusk. expanding its breeding range into areas per- Animal Behaviour 61:181–189. mi ing winter survival of part of the popula- A, T. 1978. Analysis and a theory of tion, or when climate change increases winter visible bird migration. Oikos 30:273–349. survival of resident individuals. The migra- A, T., Å. L¾. 1990. Optimal tory traits must become partially suppressed bird migration: The relative importance in such cases, but we can expect the same of time, energy and safety. Pages 331–351 type of weather response among individuals in Bird Migration: Physiology and migrating in such a population as among the Ecophysiology (E. Gwinner, Ed.). Springer- ancestral regular migrants. A possible outcome Verlag, Berlin. of the evolution from migration to partial B, P. 1996. Control of Bird Migration. migration could be the more rigid manner of Chapman and Hall, London. obligate partial migrants, where individuals B, P. 2001. Bird Migration: A General are either migratory or resident, regardless Survey. Oxford University Press, Oxford. of environmental conditions (Terrill and Able Dƒ, J., Å. L¾. 2001. Optimal 1988). One such species may be the Blackcap departure decisions of songbirds from (Sylvia atricapilla), which has become partially an experimental stopover site and the or wholly sedentary in some southerly regions signifi cance of weather. Animal Behaviour (Berthold 1996). The Blackcap would also make 62:235–243. a good within-species comparison for evaluat- Dl, H. 1996. Migration: The Biology of Life ing the weather response in relation to migra- on the Move. Oxford University Press, New tory habit in future studies. York. Diff erences in ancestral conditions and evo- E, S. T. 1975. Migration: Orientation and lutionary processes could be further elucidated, navigation. Pages 129–219 in Avian Biology, if weather responses and other essential compo- vol. 5 (D. S. Farner, J. R. King, and K. C. nents of migration were investigated in other Parkes, Eds.). Academic Press, New York. partial migrants. Gzz}, S. A., J., K. P. A. 1970. Wind and the direction of nocturnal A|l songbird migration. Nature 228:476–477. Hl, A. J. 1996. Genetic basis, mode of We are very grateful to G. Roos and N. inheritance and evolutionary changes of Kjellén, who performed the observation counts migratory directions in palearctic warblers at Falsterbo and provided us with material; and (Aves: Sylviidae). Journal of Experimental to L. Karlsson, K. Persson, G. Walinder, and S. Biology 199:49–55. Ehnbom, among many others, who contributed Hol, H., L. K. 1997. Autumn to the work at Falsterbo Bird Observatory and migration of Blue Tits Parus caeruleus at kindly provided us with the ringing material. Falsterbo, Sweden 1980–94: Population We also thank two anonymous reviewers for changes, migration pa erns and recovery valuable comments. Financial support was analysis. Ornis Svecica 7:149–167. April 2006] Weather Response in Partial Migrants 547

Kl, P., F. R. M. 1989. R, G. 1984. Migration, wintering and Atmospheric structure and avian migration. longevity of birds ringed at Falsterbo (1947– Pages 109–142 in Current Ornithology, vol. 1980) [In Swedish, with English summary]. 6 (D. M. Power, Ed.). Plenum Press, New Anser Supplement, no. 13. York. S, H. G., J-Å.N. 1987. Intraspecifi c K, G. 1957. Experiments on bird variation in migratory pa ern of a partial orientation and their interpretation. Ibis 99: migrant, the Blue Tit (Parus caeruleus): An 196–227. evaluation of diff erent hypotheses. Auk 104: Kzl, C., T. F, S. J. 109–115. 1996. Impaired predator evasion in fat T, S. B., K. P. A. 1988. Bird Blackcaps (Sylvia atricapilla). Proceedings of migration terminology. Auk 105:205–206 the Royal Society of London, Series B 263: W, W., R. W. 1972. 1671–1675. Magnetic compass of European Robins. Ll, H., G. R. 1980. The infl uence Science 176:62–64. of weather on the autumn migration of tits, W, W., R. W. 1975. The especially the Blue Tit Parus caeruleus at interaction of stars and the magnetic fi eld Falsterbo, south Sweden [In Swedish, with in the orientation system of night migrating English summary]. Anser 19:1–10. birds. II. Spring experiments with European Pz, F., P. B, A. J. { N|ê . Robins (Erithacus rubecula). Zeitschri für 1996. Frequency of migrants and migratory Tierpsychologie 39:265–282. activity are genetically correlated in a bird W, R., W. W. 1995. population: Evolutionary implications. Magnetic Orientation in Animals. Springer- Proceedings of the National Academy of Verlag, Berlin. Sciences USA 93:14642–14647. W, R. 1974. Der Herbstdurchzug von R, W. J. 1978. Timing and amount Tannenmeise, Blaumeise und Kohlmeise of bird migration in relation to weather: A (Parus ater, caeruleus und major) auf dem review. Oikos 30:224–272. Col de Bretolet (Wallis). Ornithologische R, W. J. 1990. Timing of bird Beobachter 71:135–152. migration in relation to weather: Updated review. Pages 78–101 in Bird Migration: Ecology and Ecophysiology (E. Gwinner, Ed.). Springer-Verlag, Berlin. Associate Editor: R. L. Holberton