Chapter 9 Field methods
9.1 Observations: visual, moonwatching, IR, radar
Visual observations or observations by infrared cameras or radar result in counts of birds at a given time at a given location. Often, some behavioral observations are also recorded, such as flight altitude, flight direction or the behaviour of a bird resting on the ground. The different observation techniques differ in their detection ranges and in the information they provide (Table 9.1).
| Method | Detection.range | Influencing.factors | Information |
|---|---|---|---|
| visual | max. app. 500m | observer, weather, scope, distance, species | numbers, species, behaviour |
| moon | max. app. 1.5km | observer, weather, scope, distance, species | numbers (MTR), flight directions |
| IR | app. 10km | weather, distance species | numbers (MTR), flight directions |
| radar | over 10km | weather, distance species | numbers (MTR), flight directions, flight altitudes, wing beat pattern |
Example studies of visual observations:
- EuroBird Portal: Phenology of migratory species based on records of amateur ornithologists, link.
- Description of migration pattern of Marsh warbler Acrocephalus palustris to southern Africa (Dowsett-Lemaire and Dowsett 1987).
Example studies of moon watching:
- Broad front migration and the Alps as an obstacle (F. Liechti et al. 1996).
Example studies of infrared observations:
- Quantification of reverse migration in southern Sweden (Zehnder et al. 2002).
Example radar studies:
Flight directions of migrating passerines along the autumn migration route (F. Liechti, Komenda-Zehnder, and Bruderer 2012).
Passerines migrating across the Sahara fly at night and land for the day (Schmaljohann, Liechti, and Bruderer 2007).
9.2 Marking and reencountering
Marking birds in order to find out how far and where they move has been started by the Danish Hans Christian Cornelius Mortensen (1856 – 1921). In 1899 he, for the first time, marked Starlings Sturnus vulgaris with aluminium rings that were uniquely numbered and contained an address to make a potential finder reporting the ring (Mortensen 1901). Motivated by the success of Mortensen, the first ringing station was initiated in Rossiten (today Rybachi) on the Kurische Nehrung in 1900. Shortly after, other ringing stations were started on Helgoland and in the UK.
Today, around 5 million birds are ringed annually at many ringing stations by professional and private ringers in Europe (Balmer et al. 2008). The ring numbers are coordinated within the countries by the governments who normally delegate this job to ringing schemes. The European ringing schemes are connected within the EURING society. In regular meetings and workshops, methodological standards are developed and the EURING data base, a data pool of all marking and reencounter data, maintained. EURING also started the EURING Analytical meeting that serves as a network of scientists developing statistical and mathematical tools for the analysis of mark-reencounter data.
The capturing and marking methods underly national policies. In Switzerland, only trained persons can apply for a permission for capturing and marking birds.
Mark-reencounter data is widely used to estimate survival, population sizes, movement pattern, migratory connectivity, and stop-over durations. Du Feu et al. (2016) review the scientific achievements of bird ringing, present the aims and policies of the EURING data base and give perspectives of bird ringing.
A central challenge in the interpretation of mark-reencounter data is the spatial and temporal heterogeneity of reencounter probability (Perdeck 1977; Fränzi Korner-Nievergelt et al. 2010; Thorup et al. 2014). Mark-recapture modelling techniques such as the Cormack-Jolly-Seber (CJS) model (Cormack 1964; Jolly 1965; Seber 1965), the multi-state model (Arnason 1972; Arnason, Schwarz, and Gerrard 1991) or mark-dead recovery model (Brownie et al. 1985), are seminal techniques that are nowadays widely applied and integrated in a variety of different ecological models. These models estimate reencounter probability and thereby account for its heterogeneity while interpreting mark-reencounter data.
Many countries published their mark reencounters in migration atlases, e.g. Bønløkke et al. (2006), and Saurola, Valkama, and Velmala (2013).
Examples of quantifying migratory connectivity studies based on ring reencounters:
- Estimation of the proportions of some song bird species migrating to Africa (Thorup and Conn 2009).
- Origin of Woodcocks hunted in France (Bauthian et al. 2007).
- Migratory connectivity of European Nightingales Luscinia megarhynchos (Fränzi Korner-Nievergelt, Liechti, and Hahn 2012)
- Migratory connectivity of European reed warblers Acrocephalus scirpaceus (Petr Procházka et al. 2017).
For estimating survival, it makes a fundamental difference whether the marked animal is found dead or whether it is resighted or recaptured alive. When it is found dead the time of death and thus the lifespan of the individual is known. If we assume that the probability that a dead animal is found and its mark reported is independent of its age, then annual survival estimates are straight foreward. For example, using dead recoveries of 8 different songbird species ringed in Denmark, Lerche-Jørgensen et al. (2018) found that early arriving long-distance migrants in spring have a lower survival compared to late arriving individuals whereas arrival date was not correlated with survival in short-distance migrants. Interestingly, female long distance migrants arrived in average at a date correlated with highest survival, whereas males arrived earlier thereby risking an increase in mortality.
When marked animals are recaptured and released alive or resighted, the exact time of death is not known but a minimum life-span is known. In order to estimates survival from such data, it is necessary to estimate recapture or resighting probability (reencounter probability). If reencounter probability is known we can estimate the probability that an individual is still alive dependent on the time after its last reencounter. The CJS model allows estimating both reencounter and (apparent) survival probability. However, because both dead individuals and individuals that emigrated from the study area can no longer be reencountered, the model cannot distinguish between emigration and death. Therefore, the “survival” estimates from mark-recapture/resighting data have to be interpreted as the probability that an individual stays in the study area and survives, the so called apparent survival.
Example studies:
- Sillett and Holmes (2002) estimated lower apparent survival during migration compared to the breeding and non-breeding periods in the Black-throated blue warbler Dendroica caerulescens.
- In the central Apennine, female White-winged snowfinches Montifringilla nivalis disappear at a higher rate in years with warm and dry summers, whereas the apparent survival of males does not seem to depend on summer weather (Strinella et al. 2020).
9.3 Tracking individuals
The vision of being able to follow a migratory bird on a computer screen has become, at least for larger animals, reality by now. Geographic position, flight altitude and other variables can be measured within short time and related to individual characteristics.
Two major principles exist for tacking devices. Archival tags store information on geographic position or anything else locally and the tracking device needs to be retrieved to access this information. In contrast, transmitters actively transmit information to a receiver such that data can be remotely downloaded (Table 9.2). The development of tracking devices is ongoing intensively. Devices become smaller, energy households more efficient, type of measurements more diverse and programming more flexible.
| Tag type | Data | Pros and cons |
|---|---|---|
| Archival tags | ||
| Geolocators | Light and time | +:light weight (0.5 - 1.5g), recapturing required; -:data analyses complex, location estimate with large error range (equinox) |
| Multi-sensor loggers | light and time temperature pressure activity | +:additional information such as flight height, activity, etc., combination of data for better location estimation; -:recapturing necessary |
| GPS loggers | GPS position temperature pressure activity salinity | +:huge amount of information, precise location information; -:recapturing necessary, relatively heavy |
| Transmitters | ||
| PTT tags, RFID Radio frequency identification | identification of an individual with a reader at a strategic place (nest, feeding site) | +:low cost, small and leight-weigt (sometimes implanted); -:readers are expensive, individual is registered only when it is close to the reader (data are often noisy) |
| Radio telemetry, VHF very high frequency telemetry | identification of an individual by an antenna over short to medium long distances (0-ca. 20km) | +:remote data collection, can be very light (0.3g), precise location information; -:battery life limited for small transmitters, high effort for data collection |
| Satellite telemetry, e.g. ARGOS | GPS position of individuals at any time anywhere on the globe | +:remote data collection via satellite over very long distances, almost no effort needed for data collection, precise location information; -:large weight, expensive |
The Motus project is a world wide radio telemetry collaboration. Antennas around the world register radio tagged animals and send the information to the researchers.
The movebank data base is collecting tracking data from different studies. Its aim is to foster collaboration between different researchers. Some of the data are even provided for free.
Example studies using geolocation:
- Non-breeding areas of Barn swallows Hirundo rustica (Felix Liechti et al. 2014).
- Migration pattern in relation to ecological barriers: (Hahn et al. 2014).
-Effects of geolocation on behaviour and body condition: Geolocators that weighted more than 2.3% of the body mass of tracked bird negatively influenced return rates in waders (Weiser et al. 2016).
Example studies using radio telemetry:
- Bächler and Schaub (2006) estimated stop-over durations in an oasis of the Sahara using mark-resighting data and compared these estimates with stop-over durations measured using radio telemetry.
- At an Alaskan stop-over site,Schmaljohann et al. (2013) studied the effect of fuel load and weather on stop-over duration of the Wheatear Oenanthe oenanthe.
- The first animal that was tracked by radio telemetry was a Grizzly bear Ursus horribilis (Craighead and Craighead 1972).
Example studies of satellite telemetry:
- Kjellén, Hake, and Alerstam (1997) tracked two Ospreys Pandion haliaetus from the same breeding location. The two individuals used completely different migration routes. The technique is nowadays widely applied for the study of movement behaviour of larger birds.
9.4 Orientation experiments
Kramer (1949) detected that migratory birds show a directional preference. This behaviour has widely been used to study bird orientation using orientation cages. Emlen and Emlen (1966) designed an orientation cage that is still in use, the so-called “Emlen-funnel”. Originally, ink was used to visualize the birds tracks. The ink was replaced by typewriter paper on which birds leave scratches while their feathers stayed clean. Some methodological issues with this type of orientation cage are discussed by Nievergelt and Liechti (2000). After typewriter paper has no longer been produced it was recently replaced by thermal paper (H. Mouritsen et al. 2009). Automatic registration of bird activity reduces the amount of work for data collection, e.g. Henrik Mouritsen and Larsen (2001) or Beck and Wiltschko (1983).
9.5 Stable isotopes, genetics and physiology
Stable isotopes indicate humidity and reflect trophic levels of the diet during the time when the bird’s feathers grew. In America, the isotopic map shows a north-south gradient (Hobson and Wassenaar 2008). Hobson et al. (2012) developed a isotopic map for Africa that may serve as a basis for studying bird migration and non-breeding ecology. However, geographic information in isotopes is weak. Isotopes measure chemical composition of the diet and may better be interpreted as such (Hahn et al. 2013). For extracting geographic information, often additional data is needed. For example, P. Procházka et al. (2013) combined ring reencounters with isotopes to describe a migratory divide in the Reed warbler Acrocephalus scirpaceus.
When populations differ genetically, the DNA of individuals at stop over or non-breeding sites can reveal their breeding origins. The analysis of mtDNA showed that Dunlins Calidris alpina observed in Portugal during migration originated from Greenland, Iceland and the Baltic Sea whereas the ones staying over winter had haplotypes of populations further to the east (Lopes, Marques, and Wennerberg 2006).
The study of metabolites in blood samples revealed that birds use lipids and proteins as energy fuel for endurance flights. In a review, Jenni-Eiermann and Jenni (2012) describe how the proportion of protein used is related to the duration of fastening, such as during a long migratory flight.