Observing Migrating Butterflies

These instructions are for observers who want to make detailed and extended observations of the flight directions of migrating butterflies.

More casual observers reporting observations of migrating butterflies for the map do not necessarily need to do as much work as is described here.  Casual observers are often surprised by seeing butterflies migrating and then make a quick report of the direction toward which they are flying (perhaps with reference to streets or other landmarks) and sometimes estimate how many they see in perhaps five or ten minutes.  Such reports are very useful to find out where butterflies are migrating and where they are heading.  When combined with similar reports from other observers in your area, even casual sightings can show migration patterns in interesting and surprising detail.

With that said, here are some instructions for those more dedicated observers who like to watch and study migrations for longer periods of time:

What equipment do I need to observe migrating butterflies?

Basic equipment includes a pen or pencil, a notebook or tape recorder, a compass, an accurate watch, and binoculars. If you can find binoculars with a built-in compass, these can be useful for observing any side-to-side oscillation in flight direction. You will also need information on the weather conditions, including air temperature, wind velocity, wind direction, and either cloud conditions (type of clouds and percent of sky covered) or solar radiation. The local weather station is a good source of hourly weather data. You may want to record weather conditions yourself, especially if you are interested in the effects of shorter-term changes in the weather. You can find more information on recording detailed weather data in the Notes on Observing Methods later in this section.*

Where should I go to watch migrating butterflies?

The best sites for observing migrating Vanessa butterflies are large open areas (Fig. 2) as far as possible from trees or buildings or from landforms such as hills, ridges, mountains, ravines, and arroyos, which may cause butterflies to deviate from their prevailing flight direction.** Find an open area through which Red Admirals or Painted Ladies are migrating and choose a spot from which you can easily observe them.

Figure 1. Rose map shows flight directions of Red Admirals in northern Italy during a fall migration (Benvenuti et al., 1994). Procedures described here are based upon the methods of Benvenuti et al.(1994, 1996) and Schmidt-Koenig (1979).

Figure 2. Migration observing site in Ames, Iowa, summer 2001. This view is toward the north. From this location we observed the great 2001 Red Admiral northward spring migrations.

How should I record the flight direction of each butterfly?

Choose butterflies which are flying in a consistent direction rather than just circling or moving randomly about.*** If a butterfly passes directly over you, follow it with binoculars until it vanishes from sight. Then record its bearing with the compass.**** Set the compass due north and determine to the nearest degree the angle from due north that the butterfly is flying. 0 degrees is north; 90 degrees, east; 180 degrees, south; 270 degrees, west, etc. Whenever a butterfly does not pass overhead, but close enough to you, run as fast as you can to the point where the butterfly was first spotted, then record its vanishing bearing in the same way. This procedure minimizes parallax bias which can result from trying to observe from off to one side. If you have trouble accurately sighting the direction toward which a butterfly is going, locate a distant object in the same direction and point the compass at it instead. If you are working on sloping areas or on sites with loose or stony ground, avoid running across them and follow only butterflies which you can reach easily.

How should I record the time of each observation?

Because previous observers have found migration direction to change with time of day (Baker, 1978, p. 434) or with the temperature at particular times of day (Benvenuti et al., 1996), the time of each observation should be recorded, at least to within 5 minutes. To the nearest minute is better. You should set your watch accurately, at least to the nearest minute. For some research, to the nearest second may be better. Accurate time signals are broadcast from the U.S. National Institute of Standards and Technology radio station WWV, Fort Collins, CO. Carrier frequencies are 2.5, 5, 10, 15, and 20 MHz. WWV gives Coordinated Universal Time (UTC). To convert to U.S. time zones, subtract 5 hours for Eastern Standard Time, 6 for Central, 7 for Mountain, 8 for Pacific, 10 for Alaska-Hawaiian. If you are observing when Daylight Saving Time is in effect, be sure to state whether you are recording your observations in Daylight Saving Time or standard time. States such as Arizona and Hawaii use standard time year-round.

How long is a typical observing period? What are the best times of day to observe migrating butterflies?

Observing sessions typically last from one up to six hours (if you are especially patient or enthusiastic). Because butterflies need sunshine to stay warm enough to fly vigorously, migration is best observed several hours either side of early afternoon. Chip Taylor of Monarch Watch recommends two-hour periods at these times: 10 a.m.-12 n., 12 n.-2 p.m., and 2 -4 p.m. Do not observe for such a long time that you get tired or lose interest. Doing so can introduce bias into the data. Longer observing periods are best divided among several observers, especially if observing continues for many days.

Is it better to observe butterflies over an extended period of a few days to several weeks?

Yes. To thoroughly understand the pattern of migration, you should observe butterflies for at least several days at different times and in different weather conditions. Abundance, flight speed, and flight direction of migrating butterflies often depend on both time of day and weather conditions such as temperature, wind speed and direction, and cloud cover. On cooler days, butterflies may be able to migrate only around mid-day. V. atalanta butterflies of some populations may change their flight direction as the sun' s azimuth (compass bearing) changes (Baker, 1978, p. 434). Shareware and freeware programs such as SunPo and Home Planet can be used to calculate solar azimuth for such purposes.

How precisely do I need to know where my observing site is located?

You'll need to know your location fairly accurately. You can locate your site precisely with USGS topographical maps. You can often find these at a local library or municipal government offices. If at all possible, locate your site to the nearest mile or nearest minute of arc in latitude and longitude. You can also give your location with respect to known landmarks or numbered north-south and east-west roads, or as section, range, and township numbers within a county. Best of all is a GPS (Global Positioning System) indicator. With one of these, you can locate your site to arc-second (or better) accuracy. You can get similar accuracy from a mapping site such as ACME Mapper.

How do I plot and analyze flight direction data?

Flight direction data are typically plotted on a radial "rose graph", with lines or arrows extending outward from the center of the graph (see top of page). The angular position of the line or arrow corresponds to flight direction. The length of each line or arrow is proportional to the number of butterflies seen flying in that direction. If there is a preferred flight direction, the arrows pointing nearer to that direction will be longer. If you understand statistics, you can use circular statistics (Mardia, 1972; Batschelet, 1981) to determine the mean flight direction, the variance in flight direction, and whether butterflies tend to fly significantly more often toward the mean flight direction. You can also find out if the mean flight directions in two or more sets of data taken at different times or in different conditions are significantly different.

Software for computing circular statistics is also available. The program Oriana calculates statistical parameters and plots circular graphs in several different formats.

What else besides flight direction can I record while observing migrating butterflies?

Some other possible observations include hourly and daily counts of migrants and flight speed.

How do I count the number of butterflies passing through my observing site?

You can record the number of Red Admirals or Painted Ladies you sighted each hour during the observing session. If too many butterflies are passing by, however, some may pass unseen while you are measuring the flight directions of others. If you want to do both at once, it is better to be one of a pair or group of observers. Only butterflies which are within 30 to 50 meters away (depending upon the color of the background) will typically be visible. For this reason, it is a good idea to define a line of known length, perhaps 30 meters or 100 feet long, that is perpendicular to the general direction of flight, then record the number of butterflies crossing this line per minute, per hour, or per some other interval. Because butterfly migration rates often vary widely, from a few to thousands of individuals per hour, the best time interval to report results may likewise vary. For example, Derham Giuliani, who observes enormous Painted Lady migrations through California, typically records his results as number of butterflies/50 feet/5 minutes.

If several observers count butterflies in the same place at the same time, they should stay at least 100 m apart so that no butterfly will be counted by more than one person. You should observe long enough to get enough data to calculate an accurate hourly rate. This is especially true if butterflies are arriving in bunches separated by intervals when none are passing by. You can also accumulate your total counts for each day over a period of days to estimate the start, end, peak day, and duration of the migration. If you are doing so, I recommend taking data during the same period of time each day.

How do I measure the flight speed of a migrating butterfly?

Flight speed can be determined by measuring the time required for single butterflies to move between two poles placed at least 40 m apart along the prevailing flight direction. For precise results, the butterflies chosen for observation should be those flying as near and as parallel to both poles as possible.

What is a good way to plot hourly and daily counts and flight speed of migrants?

Such observations can be plotted as bar charts. Number of butterflies seen each hour or day is plotted on the vertical axis and hour or day on the horizontal axis. Distribution of flight speeds can be plotted with speed on the horizontal axis and the number of butterflies flying at a particular speed on the vertical axis. You can also try plotting counts or mean flight speeds against wind direction, wind speed, temperature, time of day, etc.

Additional Notes on Methods

How should I record weather conditions?

If you measure weather conditions yourself, you should do so at least hourly, and perhaps more often when conditions change frequently. You can record temperature with an ordinary thermometer. This should be shaded (a piece of cardboard or a toilet tissue core painted gloss white surrounding the thermometer bulb is good) and mounted on a pole at a consistent height above the ground out in the open. This can be one of the standard heights such as 5 feet, 1.6 m, or 2 m used by meteorological stations, or you can estimate and use the average altitude of the migrating butterflies. More frequent and precise measurements can be obtained with more precise thermometers, or a shaded thermocouple connected to a digital thermometer or automated datalogger. When I use a datalogger, I usually take readings once per second and average these once per minute.

A simple homemade wind vane may show the wind direction accurately enough for some purposes; wind vanes of the type used with a datalogger are much more precise. Wind velocity can be roughly estimated by observing the movement of various objects in the wind (Beaufort scale); such estimates should be compared with wind velocity measured at a local weather station. Wind velocity can be measured more precisely with a hand-held or pole-mounted anemometer. Solar radiation can be measured with a pyranometer (watts/square meter) or a light meter. This should be either kept level or pointed directly at the sun during each reading. An alternate method is to note the type and extent of cloud cover present (e.g., clear, cirroform, cumuliform, overcast) during the observing period. On days with broken cloudiness, the times when the sun goes in and comes out can be recorded. Some correlation of the sunny periods with increased migration rate may be evident.

How might objects and landforms affect migratory movements of butterflies?

Red Admirals and Painted Ladies often fly toward hilltops and along ridgetops, trails, cliffs, and similar visually prominent features. If you are specifically interested in finding out how butterflies fly around or along such features, this too can be an interesting study. However, such data should not be lumped together with observations of migration direction in open areas.

Why are the movements of butterflies sometimes non-directional, even on days when they have been migrating?

If you see Vanessa butterflies circling or milling about during the late afternoon, they may be starting perching or territorial behavior, and they will stop migrating for the rest of the day. Vanessa butterflies seen in early to mid-morning may have just left their treetop roosts and may bask and circle to warm themselves up.

Should I record flight headings as geographical or magnetic directions?

Magnetic north as shown by a compass varies from true north by several degrees in most places. You should state whether your angular measurements are recorded with reference to magnetic or true north. The difference between the two is called magnetic declination. To find magnetic declination for your area, you'll need to know two things: your location and the date when you observed. You'll need the date because magnetic declination changes gradually over time as the earth's magnetic field varies. One site where you can find this information is the Historical Magnetic Declination site, maintained by the (U.S.) National Centers for Environmental Information (NCEI).


If you would like to help out with this research or are doing similar research, please contact me. If you would like to send results (data, summaries, or reprints are O.K.) these would be appreciated and you will of course receive full credit for any information you supply. Questions or suggestions for improving these procedures are also welcome. Additional information on methods for observing butterfly migrations is available at the Tactics and Vectors website. Thank you for your interest!

Literature Cited

  1. Baker, R.R. (1978). The Evolutionary Ecology of Animal Migration. Hodder and Stoughton, London, 1012 pp.
  2. Batchelet, E. (1981). Circular Statistics in Biology. Academic Press, London.
  3. Benvenuti, S, Dall'antonia, P, and P. Ioalè. (1994). Migration pattern of the red admiral, Vanessa atalanta L. (Lepidoptera, Nymphalidae), in Italy. Bollettino di Zoologia 61: 343-351.
  4. Benvenuti, S., Dall'antonia, P. and P. Ioalè. (1996). Directional preferences in the autumn migration of the Red Admiral (Vanessa atalanta). Ethology 102: 177-186.
  5. Mardia, K.V. (1972). Statistics of Directional Data. Academic Press, London and New York.
  6. Schmidt-Koenig, K. (1979). Directions of migrating monarch butterflies (Danaus plexippus; Danaidae; Lepidoptera) in some parts of the eastern United States. Behav. Process. 4: 73-78.
  7. Scott, J.A. (1992). Direction of spring migration of Vanessa cardui (Nymphalidae) in Colorado. Journal of Research on the Lepidoptera 31: 16-23.