Ecological Archives E092-157-D1

Georg F. J. Armbruster, Manuel Schweizer, and Deborah R. Vogt. 2011. A database on visible diurnal spring migration of birds (Central Europe: Lake Constance). Ecology 92:1865.


INTRODUCTION

Migration of birds is an exciting phenomenon in animal behavior. Millions of birds migrate from breeding grounds to over-wintering areas and back again every year. Birds have thereby caught the interest of amateurs and ornithologists alike so that arrival and departure of various species at specific sites have been recorded for almost two centuries or longer (Lehikoinen et al. 2004: p. 3). The intensity of migration for a given species depends on several factors, e.g. on population dynamics, breeding success, climate, weather conditions, winter survival, and on the locality where the migration census is monitored. Across the European mainland, visible diurnal migration of offspring and parental birds is well documented in late summer and fall, and several counting sites are under systematic surveillance (e.g. coast lines or topographic 'hot spots' like mountain passes). The intensity of back migration from over-wintering sites in spring, however, is rather poorly documented with only few systematic studies available (e.g. LWVT/SOVON 2002). This is probably because of the overall reduced population size due to mortality at over-wintering sites and during migration (e.g. LWVT/SOVON 2002; Armbruster et al. 2005). Here, we present an online archive on the intensity of spring migration. The data represent a comprehensive, fine-grained set of records from an observation site at Lake Constance (Bodensee; southern Germany). Bird census was done in three consecutive years in the mid-eighties, i.e. more than 25 years ago. Our database will allow interested scientists to conduct further analyses, e.g. i) on daily migration patterns, ii) on weather conditions and migration, and on iii) species comparison in migration behavior. Furthermore, the database might also provide comparative evidence for methodologically similar studies conducted in the future. Results of our field study have been summarized in Armbruster et al. (2005).

METADATA

CLASS I. DATA SET DESCRIPTORS

A. Data set identity: Date and Year and Time, Scientific name in Latin, Migration Heading, Number, Wind and Wind Force, Precipitation, Temperature, Air Pressure, Visibility at the tip of the Höri peninsula at Lake Constance, and number of field ornithologists at the counting site.

B. Data set identification code:

C. Data set description

Principal Investigator: Georg F. J. Armbruster, University of Basel, Department of Environmental Sciences, Section of Plant Ecology, Schönbeinstrasse 6; Switzerland. Queries regarding the data set can be directed to: g.armbruster@unibas.ch

Abstract: Millions of birds migrate each year between breeding grounds and overwintering areas. Migration behavior is influenced by a number of factors: day length, physiological and genetic factors, the magnetic compass, population dynamics, and by weather and climatic conditions. Historical and current data on avian migration are of broad scientific value. Most of the information on avian migration is presented in the form of summary statistics (e.g., duration of migration, number of individuals, trend line over years) in published studies, but original data sets are not readily available. Here, we describe the first open-source archive on visible diurnal spring migration in Central Europe. The database comprises the records of a systematic field study, which covers 35541 records of migrating birds for more than 138000 specimens of 96 species. Data were collected during the spring periods of 1984, 1985, and 1986. Bird census was conducted at the tip of the Höri peninsula in Lake Constance, Germany (north of the Alps: 47°41'49.11" N, 9°00'21.52" E). Migrating birds were counted from the beginning of March to the beginning of May each morning using fine-grained 15-min intervals. Flock size, migration heading, and weather conditions are provided for each species, e.g., for the European Chaffinch Fringilla coelebs. In all three years we found evidence that Chaffinches ceased to migrate to northeastern and eastern directions at the end of the observation period. The migration pattern of two thrush species (Turdus merula and T. viscivorus) is also presented in more detail. Turdus merula had its migration peak always early in the morning, whereas migration of T. viscivorus was recorded from the early morning onward, with a peak at late morning to noon. Migration records are also documented for species that are generally believed as resident birds in the inland of Central Europe (e.g., Streptopelia decaocto, Parus palustris, Certhia brachydactyla, Pica pica). Our archive might be of significance for studying patterns of daily migration, e.g., the effect of weather on spring migration intensity and for methodological studies on the same issue.

D. Key words: bird migration, Bodensee, Höri peninsula, Lake Constance, spring avian migration.

CLASS II. RESEARCH ORIGIN DESCRIPTORS

A. Overall project description

Identity: Date and Year and Time, Scientific name in Latin, Migration Heading, Record of solitary individuals or bird flock in the time slot, Wind and Wind Force, Precipitation, Snow cover; Temperature, Air Pressure, Visibility at the tip of the Höri peninsula at Lake Constance, and number of field ornithologists at the counting site.

Originator: Georg F. J. Armbruster, University of Basel, Department of Environmental Sciences, Section of Plant Ecology, Schönbeinstrasse 6; Switzerland

Period of Study: 1984–1986

Objectives: To describe patterns of visible diurnal spring migration of birds in the inland of Europe (diurnal and temporal patterns, intensity, weather data).

Abstract: same as above.

Sources of funding: Data collection, transcription, and verification have been supported by the Ornithologische Arbeitsgemeinschaft Bodensee (Konstanz, Germany) and Basler Stiftung für Biologische Forschung (Basel, Switzerland).

B. Specific subproject description

Site description: Migrating birds were counted in all three years at exactly the same place (the tip of the Höri peninsula at the open lake bank, 395m a.s.l., at 47°41'49.11" N, 9°00'21.52" E). For better view on the area, see, e.g., Google Earth (http://www.google.com/earth/index.html). Vegetation belt at the lake bank consists of reed grass Phragmites australis (CAV.) STEUDEL, and a group of deciduous trees and conifers is situated near the counting site.

Site type: The Höri peninsula in the western part of Lake Constance is up to 708 m a.s.l., and is predominately covered with forest, agricultural land and with small villages. The peninsula has a west–east orientation. Migrating birds mostly follow this W–E direction or follow the lake bank, and arrive at the tip of the peninsula in high number of individuals.

Geography: Lake Constance (the Bodensee) is 63 km long and a maximum of 14 km wide, and lies in the northern foreland of the Alps. It is a tri-national lake surrounded by Germany, Switzerland, and Austria. It is one of the biggest freshwater reservoirs of inland central Europe (water volume ca. 50 km3). The lake area was shaped during previous ice ages. The basin of the lake was excavated by huge glaciers. Surface elevation of the water is approximately 395 m a.s.l., with the lowest water level in February. Melting water of the Alps can raise the surface of the lake up to two meters in summer. The Rhine river is the outflow of the lake, and takes the water to the North Sea.

Habitat: see above.

Geology: The slopes and hills around Lake Constance are mainly formed by ground moraines and molasses. The Alps reach the basin of the lake in the very southeast, at some places with spear rocks (i.e., 'Nagelfluh' conglomerate). Phonolitic and basaltic hills (formed by volcanoes in the Miocene) are in the northwest of the basin.

Watersheds/hydrology: see above.

Site history: The observation site at the tip of the Höri peninsula is famous for watching diurnal spring migration. Local field ornithologists have known this place for many years. It is also a popular place in winter to watch water fowls, cormorants, and loons.

Climate: The climate at Lake Constance is strongly influenced by its geographic position and its huge body of water. Mean annual temperature is about 8–9 °C. In winter, the climate is relatively mild with only 1–2 weeks of strong frost. In summer, daily temperatures can reach 25–32 °C. Table 1 shows detailed information of the weather station of Konstanz (located 47°40'34.5" N, 9°11'29.4" E), ca. 12 km east of the counting site. Additional time series on weather data can be found online in the open source archive of the Deutscher Wetterdienst (http://www.dwd.de).

TABLE 1. Average minimum temperature and average maximum temperature (1961–1990; weather station of Konstanz).

  Min Temp.
(°C)
Max Temp.
(°C)
March -1 8
April 3 13
May 7 17
Coldest month (January) -5 2
Warmest month (July) 12 23

At Lake Constance, a warm southern wind named "Föhn" can blow, particularly in the southeastern part. Föhn winds sometimes take Mediterranean birds over the Alps to the lake. Precipitation in the lake basin follows a gradient from ca. 700 mm p.a. in the northwest to 1300 mm p.a. in the very southeast.

Experimental design: Data are the result of a three-year observational field study.

Design characteristics: N/A

Sampling methods: The observation site was occupied with one field ornithologist who was familiar with identifying migrating birds. In rare cases, two field ornithologists have done the census (see database). Most of the field work was done by the P.I. (GFJA). Birds were identified by eye, binoculars (8–10× magnification) and/or by typical migration calls. If possible, the migration heading of each specimen or bird flock was recorded when the birds crossed the counting point. Data were recorded in handwritten field protocols. Note that swifts (Apodidae) and swallows (Hirundinidae) do not appear in the database. This is because local individuals (e.g., hunting for insects) and actively migrating individuals could not be distinguished at the observation site.

Taxonomy and systematics: Latin names of birds follow Maumary et al. (2007). Note that the data entries of the Lesser Redpoll Carduelis cabaret may also comprise specimens of the Common Redpoll Carduelis flammea.

Permit history: A permit for visiting the tip of the Höri peninsula must be applied for at the Naturschutzbehörde Konstanz (i.e., the government department of nature conservation, county of Konstanz). The area is under protection, and access is restricted.

Legal/organizational requirements: None.

Project personnel: Georg F. J. Armbruster, Andreas Schmidt, Arne Brall, Christoph Braunberger, Rüdiger Specht, Siegfried Schuster.

CLASS III. DATA SET STATUS AND ACCESSIBILITY

A. Status

Latest update: The data spans the period of 1984–1986.

Latest Archive date: 4th May 1986

Metadata status: The metadata are complete and up to date.

Data verification: Field protocols were written into Excel files. These data entries were then checked for typing errors.

B. Accessibility

Storage location and medium: The original Excel file is stored on Apple computers in the Plant Population Biology Group of the University of Basel, Switzerland.

Contact person: Georg F. J. Armbruster, University of Basel, Department of Environmental Sciences, Section of Plant Ecology, Schönbeinstrasse 6, 4056 Basel, Switzerland; g.armbruster@unibas.ch

Copyright restrictions: None.

Proprietary restrictions: None.

Costs: None.

CLASS IV. DATA STRUCTURAL DESCRIPTORS

A. Data Set File

Identity: BirdMigrationData.txt

Size: This file contains the 35541 migration records not including header row. Each row consists of 17 cells.

Format and storage mode: ASCII text, tab delimited. No compression was used.

Header information: See variable names in Section B.

Alphanumeric attributes: Mixed.

Special characters/fields: Missing data denoted as -9.

Authentication procedures: See Section III. A.

B. Variable information

Data in the file BirdMigrationData.txt are organized laterally in rows. Each row of the data represents one bird flock or solitary individual passing the observation site. Variables are in columns.

Variables in the ASCII file (= columns in table format) are:

month: numbers indicate the month, i.e., 3 = March, 4 = April, 5 = May
day: e.g., 6 = 6th of the respective month, 31 = 31st of the respective month. The following dates are described in detail because these days do not appear in the database:
7th May 1984: missing information (see CLASS V. SUPPLEMENTAL DESCRIPTORS, B., below).
8th May 1984 and 28th April 1986: no migration activity was detected, although the counting site was occupied by a field ornithologist (see the list of observation hours per day in separate ASCII file " ObsHoursPerDay.txt").
19th March 1985: On that particular day, a 10-cm layer of fresh-fallen snow was recorded at Lake Constance (protocol of the Germany’s Meteorological Service Institute in Konstanz located 47°40'34.5" N, 9°11'29.4" E at 440 m a.s.l., ca. 12 km east of the counting site). No active spring migration was detected at the tip of the Höri peninsula although the counting site was occupied by a field ornithologist (see the list of observation hours per day in separate ASCII file " ObsHoursPerDay.txt").
year: the year of investigation, 1984, 1985, or 1986.
pent: five days interval (i.e., "Pentade"), e.g., 13 = 2nd–6th March, 17 = 22nd–26th March, 25 = 1st–5th May.
sunq: 15 minutes interval ("Quartile") adjusted to local time of sunrise. Examples: 0 = 15 min before sunrise to sunrise, 1 = from sunrise to 15 min after sunrise, 5 = one hour after sunrise to 75 min after sunrise.
taxon: scientific name of bird species. In few cases, only genera were identified, followed by "sp." (e.g., warblers of Phylloscopus sp.).
direction: migration heading of each bird flock or solitary individual at the observation site, e.g., Unknown = unknown movement heading, NW = migration heading to northwest. Description of directions:
E = heading over the lake
NE = heading over the lake
N = heading over the lake
NW = heading along the lake bank
W = heading to the mainland of the peninsula
SW, S and SE = heading to the southern bank of the peninsula, or to the Swiss lake bank
number: absolute number of specimens in the time slot, e.g., 1 = a single specimen, 32 = a flock of 32 individuals.
winddir: wind from one of the eight directions. A dash "-" indicates no wind.
windforce: is given in a relative scale from 0 to 6, i.e., 0 = no or nearly no wind, 6 = strong wind, in about 6 Beaufort.
rain: relative scale, with 0 = no precipitation, 1 = weak, some drops, 2 = continuous, 3 = strong rain fall.
snowfall: relative scale, with 0 = no snowfall, 1 = some flocks, 2 = continuous snowfall at the counting site.
snow layer: height of snow layer (in cm) at Germany’s Meteorological Service Institute in Konstanz (located 47°40'34.5" N, 9°11'29.4" E at 440 m a.s.l.), ca. 12 km east of the counting site, e.g., 0 = no snow layer, 6 = a 6-cm high layer of snow, R = rests of snow, i.e., no complete snow cover.
temperature: temperature (in °C) at 6.30h a.m. at Germany’s Meteorological Service Institute in Konstanz.
mbar: air pressure at 6.30h a.m. at Germany’s Meteorological Service Institute in Konstanz.
visib: visibility at the observation site. 0 = no mist or fog, 1 = weak mist, visibility up to 2 km, 2 = visibility below 2 km, 3 = mist or fog with visibility below 1 km.
observer(s): number of field ornithologists at the counting site. At most days, just one field ornithologist performed the census.
Note: missing values appear with -9.

A. Data Set File

Identity: ObsHoursPerDay.txt

Size: This file shows the observation hours per day in each year (195 data not including header row; each row consists of 4 cells).

Format and storage mode: ASCII text, tab delimited. No compression was used.

Header information: See variable names in Section B.

Alphanumeric attributes: Mixed.

Special characters/fields: Missing data denoted as -9.

Authentication procedures: See Section III. A.

B. Variable information

Data in the file ObsHoursPerDay.txt are organized laterally in rows. Variables are in columns. Variables in the ASCII file (= columns in table format) are:

month: 3 = March, 4 = April, 5 = May
day: e.g., 6 = 6th of the respective month, 31 = 31st of the respective month
year: in the year 1984, 1985, or 1986, respectively
obs-hours: observation time at the counting site at that day (in hours)

CLASS V. SUPPLEMENTAL DESCRIPTORS

A. Data acquisition

Data forms: field protocols on paper

Location of completed data forms: The field protocols stored by Georg F J. Armbruster, University of Basel, Plant Population Biology Group, Schönbeinstrasse 6, 4056 Basel, Switzerland.

Data entry/verification procedures: The original data were recorded by hand in field protocols. Finally, the data were transcribed and entered into an electronic MS Excel file by M. Schweizer and D. R. Vogt.

B. Quality assurance/quality control procedures: The original data entries in the Excel file were reviewed and corrected by an assistant and by the P.I. We searched for outliers, and looked for irregularities by running different sorting procedures for each species. We eventually noticed that 7th May 1984 is a "real" missing day (the observation site was not occupied by a field ornithologist).

C. Related material:

Tables 2, 3, and 4 (below) show details in the migration heading of the European Chaffinch Fringilla coelebs at the observation site in each year. E, NE, and N indicate directions across the lake, whereas other directions are more inland or lake bank directions. Two time spans are compared: beginning of March (start of the bird census) until 20th of April (denoted with March-April), and 21st of April until May (end of the bird census; denoted with EndApril-May). Chaffinches ceased to migrate in an E and NE direction at EndApril-May in each year. This finding could indicate that chaffinches heading to breeding sites in northeastern and eastern Europe may have passed the foreland of the Alps before 20th of April. Alternatively, one could speculate that migration instinct in chaffinches is already considerably reduced in late April / May, which would prevent them from flying over a large body of water towards E or NE.

TABLE 2. Migration heading of the European Chaffinch in the year 1984. Percentage of the total number of migrating individuals in one of the nine directions (with absolute number in parentheses). Rounding error can lead to values of +/- 100% per column. Based on Chi-square estimates of absolute numbers, the table cells are differently occupied between the two time periods (P < 0.001).

Direction March-April 1984 EndApril-May 1984
E 14 % (1976) 1 % (21)
NE 40 % (5790) 3 % (48)
N 3 % (420) < 1 % (3)
NW 19 % (2679) 43 % (672)
W < 1 % (74) 10 % (162)
SW 2 % (280) 14 % (216)
S 1 % (216) 6 % (99)
SE 6 % (852) 5 % (78)
Unknown movement heading 14 % (2088) 17 % (268)
Total number of individuals 14,375 1567

 

TABLE 3. Migration heading of the European Chaffinch in the year 1985. Percentage of the total number of migrating individuals in one of the nine directions (with absolute number in parentheses). Rounding error can lead to values of +/- 100% per column. Based on Chi-square estimates of absolute numbers, the table cells are differently occupied between the two time periods (P < 0.001).

Direction March-April 1985 EndApril-May 1985
E 10 % (638) < 1 % (2)
NE 13 % (834) 0 % (0)
N 4 % (283) 0 % (0)
NW 34 % (2296) 18 % (220)
W 2 % (150) 24 % (292)
SW 5 % (327) 20 % (241)
S 7 % (486) 24 % (292)
SE 11 % (760) 9 % (105)
Unknown movement heading 13 % (885) 6 % (78)
Total number of individuals 6659 1230

 

TABLE 4. Migration heading of the European Chaffinch in the year 1986. Percentage of the total number of migrating individuals in one of the nine directions (with absolute number in parentheses). Rounding error can lead to values of +/- 100% per column. Based on Chi-square estimates of absolute numbers, the table cells are differently occupied between the two time periods (P < 0.001).

Direction March-April 1986 EndApril-May 1986
E 8 % (1169) 1 % (7)
NE 37 % (5334) 2 % (10)
N 5 % (678) < 1 % (2)
NW 24 % (3539) 38 % (228)
W < 1 % (46) 8 % (48)
SW 3 % (406) 7 % (41)
S 4 % (608) 13 % (80)
SE 10 % (1404) 24 % (143)
Unknown movement heading 9 % (1284) 7 % (43)
Total number of individuals 14,468 602

 

Tables 5, 6, and 7 (below) show details in the migration behavior of two thrush species, Turdus merula (Blackbird)  and Turdus viscivorus (Mistle Thrush). In each year, about half of the total number of T. merula migrated in the first hour after sunrise, and migration ceased until late morning / noon. In T. viscivorus, however, just ca. 20% of the individuals migrated in the first hour after sunrise, and a migration peak appeared in late morning / noon. Similar results have been found in the Netherlands (LWVT/SOVON 2002: pp. 205 and 213; see diagrams "Doortrek over de dag"). An explanation for this difference could be that T. merula prefers to migrate during night and in the early morning (see Brüchle-Ulmer and Hemprich 1999), and T. viscivorus is more a diurnal migrant.

TABLE 5. Daily migration pattern (in percent of total number of individuals) of two thrush species Turdus merula and Turdus viscivorus in the year 1984 (with absolute number in parentheses). " <1h" = from start of observation until 60 min after sunrise, " >1h–2h" = from 60 min after sunrise until 120 min after sunrise, "Late Morning" =  > 120 min after sunrise until late morning / noon. Rounding error can lead to values of +/- 100% per column. Based on Chi-square estimates of absolute numbers, the two species differ significantly in their diurnal migration pattern (P < 0.001).

1984 Turdus merula Turdus viscivorus
< 1h 54 % (125) 20 % (100)
> 1h–2h 29 % (67) 19 % (93)
Late Morning 17 % (41) 60 % (295)
Total number of individuals 233 488

 

TABLE 6. Daily migration pattern (in percent of total number of individuals) of two thrush species Turdus merula and Turdus viscivorus in the year 1985 (with absolute number in parentheses). "< 1h" = from start of observation until 60 min after sunrise, " >1h–2h" = from 60 min after sunrise until 120 min after sunrise, "Late Morning" = > 120 min after sunrise until late morning / noon. Based on Chi-square estimates of absolute numbers, the two species differ significantly in their diurnal migration pattern (P < 0.001).

1985 Turdus merula Turdus viscivorus
< 1h 59 % (36) 19 % (69)
> 1h–2h 31 % (19) 15 % (54)
Late Morning 10 % (6) 66 % (244)
Total number of individuals 61 367

 

TABLE 7. Daily migration pattern (in percent of total number of individuals) of two thrush species Turdus merula and Turdus viscivorus in the year 1986 (with absolute number in parentheses). "< 1h" = from start of observation until 60 min after sunrise, " > 1h–2h" = from 60 min after sunrise until 120 min after sunrise, "Late Morning" = > 120 min after sunrise until late morning / noon. Rounding error can lead to values of +/- 100% per column. Based on Chi-square estimates of absolute numbers, the two species differ significantly in their diurnal migration pattern (P < 0.001).

1986 Turdus merula Turdus viscivorus
< 1h 59 % (65) 19 % (55)
> 1h–2h 24 % (26) 30 % (86)
Late Morning 17 % (19) 50 % (143)
Total number of individuals 110 284

 

D. Computer programs and data processing algorithms: N/A

E. Archiving: N/A

F. Publications using the data set:

Armbruster, G. F. J., D. Renz, and M. Schweizer. 2005. Eine dreijährige Feldstudie zum sichtbaren Frühjahrszug am Bodensee (Süddeutschland). Vogelwarte 43:171–178.

G. Publications using the same sites: N/A

H. History of data set usage

Data request history: N/A

Data set update history: N/A

Review history: N/A

Questions and comments from secondary users: N/A


ACKNOWLEDGMENTS

We thank the Ornithologische Arbeitsgemeinschaft Bodensee and the Basler Stiftung für Biologische Forschung for financial support. We also thank Andreas Schmidt, Arne Brall, Wolfgang Fiedler, Rüdiger Specht, Christoph Braunberger, Siegfried Schuster, and Udo von Wicht for assistance in the field. Simon Loader, Peter Stoll, Bernd Hägele, Hans-Peter Rusterholz, Daniel Bloch, and Hans van Gasteren helped in manuscript revision. Special thanks to two anonymous referees who gave valuable comments that considerably improved the manuscript. We are also indebted to Germany’s Meteorological Institute in Konstanz for providing the meteorological data.

LITERATURE CITED

Armbruster, G. F. J., D. Renz, and M. Schweizer. 2005. Eine dreijährige Feldstudie zum sichtbaren Frühjahrszug am Bodensee (Süddeutschland). Vogelwarte 43:171–178. ISSN 0049-6650.

Brüchle-Ulmer, I., and M. Hemprich. 1999. Amsel - Turdus merula. In G. Heine, H. Jacoby, H. Leuzinger, and H. Stark. Die Vögel des Bodenseegebietes. Orn. Jh. Bad.-Württ. 14/15:603–606. ISSN 0177-5456.

Maumary, L., L. Vallotton, and P. Knaus. 2007. Die Vögel der Schweiz. 1-848. Schweizerische Vogelwarte (ed.), Sempach, ISBN 9783952300626.

LWVT/SOVON (ed.) (2002). Vogeltrek over Nederland 1976–1993. 1-431. Schuyt & Co., Haarlem, ISBN 9060975669.

Lehikoinen, E., T. H. Sparks, and M. Zalakevicius. 2004. Arrival and departure dates. Pages 1–31 in A. P. Moller, W. Fiedler, and P. Berthold, editors. The Effect of Climatic Change on Birds. Advances in Ecological Research 35, Academic Press, New York, New York, USA. ISBN 0065-2504.


[Back to E000-000]