Spring Migration and Weather in Eastern Canada: a Radar Study

Spring Migration and Weather in Eastern Canada: A Radar Study

W, John Richardson*

During the past 15 years, radars have been garets, N.B. were observed from late April to used to study migration in many areas of early June in 1970 and 1971 by direct use of Europe and North America. Radars can pro high-powered nodding height finder radars. vide continuous and relatively unbiased in Eastwood ( op. cit.) has described such equip formation about the density, directions, speeds ment. and altitudes of the migrants. Unfortunately, (iii) The directions of the migrants near it is usually impossible to identify the birds other military radar sites (Fig. 2) were deter being detected. Eastwood (Radar Ornithology, mined each day and night from April 30 to Methuen, 1967) reviews the radar technique. June 9, 1971 from PPI displays showing While one can survey migration over an digitized radar data. I will discuss the tech area many miles wide by using a single radar, nique and its limitations in detail elsewhere. it is possible to obtain an even broader view In summary, it is not possible to estimate the by using radars at several sites simultaneously. density of migration from such digitized dis During the spring in 1970 and 1971 I col plays. However, the directions of samples of lected migration data from 14 radar sites in the birds near the sites can often be deter Nova Scotia, New Brunswick, Quebec and mined. At night strong echoes from birds northern Ontario. While all 14 sites were not flying in flocks ( eg., waterfowl and shore- used simultaneously, for six weeks in 1971 birds) are over-represented in the samples. I used 12 sites. These birds often fly in somewhat different This report describes briefly the patterns of directions from the passerines, and so the spring migration in the area of study. It then nocturnal digitized data are often direction explores the relationships between the density ally biased. During the day almost all bird and direction of migration and the synoptic echoes are quite strong, and I have found no weather situation. Later publications will ex bias in the diurnal directional samples. amine these and related data in greater detail. HOUR TO HOUR VARIATIONS METHODS AND DATA OBTAINED IN DENSITY I have used three types of information m At St Margarets, as in many other parts this report: of North America, there is usually much more (i) From late March to early June in 1970 migration at night than during the day. The and 1971 I made time-lapse films of the Plan density of migration typically increased rap Position Indicator (PPI) display of a high- idly about 45 minutes after sunset. It usually powered military surveillance radar at St Mar remains fairly constant from about one hour garets, N.B. I obtained similar films from after sunset until after midnight. Then a grad Barrington, N.S. in the spring of 1971. Fig. 2 ual decrease in density begins. Daytime migra shows the locations of these radar sites. The tion is usually densest in the morning, and direction and speed distributions of the mi decreases in the afternoon. These ‘average’ grants and the amount of migration were eval patterns can be modified by changing weather uated several times each day and night from conditions. In particular, wind shifts and the these films. I have found by simultaneous onset or cessation of rain usually produce moon-watching and radar surveillance that the modifications of the average temporal pattern. St Margarets radar gives a reliable picture of the density and direction of migration. HEIGHTS (ii) The heights of migrants near St Mar During the night most spring (and fall) *Section of Neurobiology and Behavior, Divi migrants at St Margarets fly below 3500 ft. sion of Biological Sciences, Cornell University, While accurate measurements are not possible Ithaca, N.Y. 14850. during dense movements, the modal height of nocturnal migration is almost always below

684 American Birds, August 1971 Spring Migration and Weather in Eastern Canada: A Radar Study

W, John Richardson*

684 American Birds, August 1971 2000 ft. Nevertheless, on most nights there movements that occur under optimal weather are a few birds up to 7000 ft. or more. The conditions. highest echoes are usually quite intense, and While we commonly speak of “forward” probably represent flocks of waterfowl and and “reverse” movements, there is no clear shorebirds, not passerines. . dichotomy between them, at least when one The birds climb rapidly at the start of the looks at all species at once. Over a season, night. An extreme example occurred at St some birds fly in all directions at all sites. Margarets in the early evening of May 12, 1971. At that time I filmed a bird echo as it WATER CROSSINGS climbed 5100 ft. in 19 minutes while moving Drury and Keith (Ib is, 104: 449-489, 1962) seven miles horizontally. The height distribu have recorded birds departing from Massa tion of the migrants reaches its peak by 1 / 2 chusetts across the Gulf of Maine towards or 2 hours after sunset. While comparable Nova Scotia. I frequently record birds over data are not available in the spring, in autumn the Gulf of Maine moving NE-E towards the average height decreases in the latter part western Nova Scotia. These arrivals usually of the night. become denser after midnight, when birds that Well-oriented movements frequently occur departed from New England soon after sunset under total opaque overcast ( eg., nights of begin to move into the coverage area of the April 30 and May 12, 14, 16, 21 and 22, 1971 Barrington radar. Reverse movements from at St Margarets). The heights of the migrants Nova Scotia back across the Gulf of Maine on these nights were not noticeably different to New England are also common. The water from normal. However during light rain I crossing from Cape Cod to Nova Scotia is a have occasionally recorded light nocturnal minimum of 250 miles. However, most birds movements unusually high (eg., May 25, do not start and stop right on the coasts 1971). Some of these birds are apparently during either forward or reverse flights. Hence climbing above the rain. most fly considerably further than 250 miles non-stop. DIRECTIONS At night many birds move NE from New The predominant direction of the nocturnal Brunswick out over the Gulf of St Lawrence. spring migration of passerines in the Maritime There is a similar nocturnal NE movement provinces is to the N e , ENE and even E, not from eastern Nova Scotia across the Cabot to the N. This is explained by the geography Straight to Newfoundland. By day comparable of eastern North America. The “east” coast (but less dense) movements sometimes occur. is actually on a NE-SW line, and to the south However diurnal migrants frequently hesitate of the Maritimes is the Atlantic. Very few or turn at coastlines, something which noc birds approach Nova Scotia from the south. turnal migrants almost never do. By day there Northward movements in the Maritimes are are frequent concentrations of migrants moving never as dense as the strongest NE passerine north ( or south) along the east coast of New flights. The N movements are composed pri Brunswick. My field observations have shown marily of intense and fast-moving echoes. that waterfowl, gulls and Gannets are often Such echoes are typical of waterfowl and involved in these coasting movements. shorebirds, not passerines. At night, passerines produce weak, slow-moving echoes. WEATHER AND MIGRATION In Quebec dense movements to the N occur Both the amount and the direction of mi more frequently than they do in the Mari gration are strongly affected by weather. times. However even there the predominant Figure 1 summarizes the 1970 and 1971 noc direction is east of north. The departure from turnal data from St Margarets. A similar pat a N-S axis of flight is therefore not simply the tern of densities and directions is obtained result of birds paralleling coastlines that are when the diurnal data are plotted on such a within visual range. map. Fig. 2-7 show specific examples of Radar studies have shown that “reverse” weather and the direction of diurnal migration migration is much more common than was over a wide area. once suspected. Movements to the SE, S and The diagrams show that at St Margarets SW in the spring are common throughout dense migration was especially common with eastern Canada under certain weather condi S, SW and W winds. Such winds occur on the tions. However these “reverse” movements NE, E, SE and S sides of low pressure areas are rarely as intense as the nocturnal NE and on the W, NW and N sides of highs.

Volume 25, Number 4 685 i

14: 35-68, 1971) to determine the Canad. Wildl. Serv. directions of Report migration. When there Ser. was more than one mode, location could sec not be located, no point was plotted. density density categories. The arrows on the data points indicate ondary the modes modal were shown by sents one shorter arrows. night Each of point observation. thus On evenings repre when an map appropriate for for that part of tively). the I season open, used ( half-filled a and technique moving-average (see solid circles and Richardson Gunn, respec

of 1970 weather and situations. A 1971 generalized weather at map St. was Margarets, drawn N.B. showing with different synoptic most most of the common spatial relationships among the pressure systems, actual synoptic intensity features or was direction of selected migration. without At knowledge this of location I the placed a point fronts fronts and wind direction. Each evening the location on map the generalized which best represented the position of St Margarets relative to the representing density and direction of one migration hour after sun-

686 American Birds, August 1971 Figures 2-7. Directions of diurnal migration at sites numbered on Fig. 2 are: 1—Barrington, various sites in eastern Canada and Maine, May N.S.; 2—St Margarets, N.B.; 3—Caswell,' Maine; 11-16, 1971. The length of each line is propor 4-7—Mont Apica, St Denis, Chibougamau and tional to the fraction of the birds moving in that Senneterre, Que.; 8, 9—Ramore and Moosonee, direction. Weather maps for the corresponding Ont. The large circles indicate the approximate times showing pressure systems, fronts, geos- areas within which birds were regularly detected. trophic wind directions and selected isobars are Spring data have also' been obtained from Syd superimposed. Circled pressure system locations ney and Halifax, N.S. (10, 11) and from Moisie, are off the map in the direction indicated. The Quebec and Montreal, Que. (12-14).

With these weather conditions the migration are light northerly winds there is often both was almost invariably to the N, NE or E. forward and reverse migration. This forward Fig. 7 shows a good example of such condi movement into light northerly winds usually tions over a wide area. involves strong non-passerine, echoes. With Northerly winds occur when there is a high strong northerly winds one usually finds only to the west or a -low to the east. On the aver reverse migrants. After allowing for seasonal age the migration was much less dense on differences, these observations are comparable days or nights with these conditions. Further to those made by Able (Bull. Ecol. Soc. more on these occasions the majority of the A m e r. 51: 43, 1970) in Louisiana and Georgia. birds were flying to the SE, S or SW. Figure 5 After a cold front goes by, the winds are shows an example of widespread reverse often from the NW or N. On such occasions migration. reverse migration was frequent (Fig. 1). How Reverse migration almost never occurred ever, by the time some lows reach the Mari without following (northerly) winds. In con times, the associated cold front has swept trast, forward migration did occur without counterclockwise around the depression such following (southerly) winds. When it is calm, that it extends S or even SE from the low, as it often is near the center of a high, for The wind behind these cold fronts is SW or W. ward migration may be intense. When there The left side of Fig. 1 shows that behind such

Volume 25, Number 4 687 Figure 3

fronts NE migration usually occurred and was was reverse. Dense migration did not recur sometimes intense. Fig. 4 shows a specific at StM until the evening of May 7; this was example: On May 13, 1971, the winds in the first night with southerly winds after southern Quebec were SW or W behind a May 4. Fairly intense migration occurred at front, and 1 recorded extensive NE migration. StM each night from May 7 to May 12. This The same phenomenon is seen in the NE was a prolonged period of warm fair weather spring migration of Starlings in Ontario (Rich with southerly or near calm winds ( Fig. 2, 3) . ardson and Haight, Canad. J. Zool. 48: 31-39, The passage of a cold front on May 13 re 1970). Thus we must recognize that when. the duced the volume of migration at StM. How predominant direction of migration is NE, the ever the. winds behind the front were SW or density of migration is ■ often as great after W (Fig. 4) and forward rather than reverse the passage of cold fronts as ahead of them. movement occurred. On May 14 NW winds occurred and reverse movement was recorded MIGRATION DURING MAY over a wide area (Fig. 5). As usual during To further our understanding of the impor such conditions, a few birds moved to the tance of weather factors in determining mi north upwind. By the evening approaching gration patterns, I have looked in detail at favorable conditions were already producing the data from May 1971. NE movement at the westernmost radar sites. The first intense movement of the month at On May 15 the eastward movement of the St Margarets, N.B. (henceforth StM) occurred pressure systems brought favorable winds to the evening of May 2 as an approaching low most of the sites (Fig. 6 ), and one of the produced light southerly winds. Similar con densest movements of the season occurred ditions the night of May 4 produced another at StM. dense movement. At the Quebec sites, which The weather deteriorated as a low moved were behind the low, most of the migration across the area on May 16. By the evening

688 American Birds, August 1971 Figure 4

Figure 5

Volume 25, Number 4 689 of May 17 the winds were NE at StM, and probably flying with more or less following light reverse migration occurred. Data were winds. incomplete for May 18-20. However a very CONCLUSIONS dense movement did occur at StM the night of May 19 in typical favorable weather con In this report I have described some aspects ditions. Thereafter dense migration did not of spring migration in eastern Canada. I have recur at StM until May 26. During parts of emphasized that weather affects not only the the interim, reverse migration occurred with numbers but also the directions and some 1 northerly winds at most of the sites. Then times the heights of the migrants. I have tried 1 from May 26 to May 30 there were five suc to point out the usefulness of obtaining data c essive nights at StM with calm or S, SW or simultaneously from widely-separated points. W winds. On each of these five nights there All of the information presented here was were above-normal densities of migration to obtained with radars that were being used the N and NE. While more detailed analysis concurrently for aircraft control and sur is necessary, it is clear that the supply of veillance. Large numbers of air surveillance birds ready to migrate is not so depleted after and weather radars are operating in North one or two nights of good weather that dense America. Groups of the surveillance radars migration cannot continue on subsequent feed their data to centralized FAA and mili nights. tary control centres. At these control centres one can observe the displays of several radars Finally, on the last day of the month there located over an area hundreds of miles wide. was widespread reverse migration over Que Hence if more ornithologists were interested, bec and the Maritimes. On this day northerly data similar to those presented here could be winds were produced over the whole area by obtained on a much greater scale than is a low east of Labrador and a high over presently being done. Ontario. At StM there was both N and SSW movement the night of May 31. The distribu ACKNOWLEDGMENTS tion of heights was unusual, in that there were I thank the personnel of Canadian Forces two layers of birds with few in between. Stations St Margarets and Barrington for their Time-lapse films of the height-finder display invaluable assistance. S. T. Emlen, W. W. H. showed that the birds in the layer below 3000 Gunn and R. E. Muller read and criticized ft. were moving 5 5 ^ , while those forming the an earlier draft of the manuscript. This study upper layer at 5-8000 ft. were moving N. The was financed by the National Research Coun surface winds were from the north while the cil of Canada through its Associate Committee winds aloft were SW. Thus all the birds were on Bird Hazards to Aircraft.

Dates for 1971-72 Christmas Bird Counts: Saturday, December 18, 1971 to Sunday, January 2, 1972 inclusive. (Three more days than 1970-71.)

690 American Birds, August 1971