Understanding the processes that maintain the distribution and abundance of different species has been an issue of fundamental importance for biologists for centuries. It is one of the most important questions that were given a theoretical framework in the theory of natural selection by Darwin and Wallace. Yet, however, much remains to be discovered before we know whether the distribution of species across areas is a random or non-random process. That is, do the populations at a site consist of all those that happened to arrive there, or only of those with properties that allow their co-existence? Moreover, this question is at the root of research programmes dealing with biodiversity at the species level and higher levels too.

Theoretical arguments and observations suggest that non-random processes govern the distribution of many species. Thus, inter-specific competition, predation and recruitment processes may determine the number of species. Also, disturbance other than predation has been proposed as another potential force in maintaining species diversity. It should be noted that the ability to colonize new areas may be a crucial factor in the distribution of species, and that disturbance and colonization are not always independent because the former often precedes the latter.

There are, however, examples of colonization events, natural ones, where no previous environmental change in the colonized area occurred. It may not be surprising that in the majority of these events, there have been birds that have successfully occupied, survived and reproduced in new areas. Moreover, a random process such as unpredictable weather during the spring migration seems to have caused the colonization event on some of these occasions.

Most birds possess excellent equipment for moving over large areas in contrast to animals that spread by wind and ocean currents. Birds have a high degree of aerial agility which enables them to cross hostile environments on their way to and from breeding grounds and, as is apparently the case for some avian lemming predators, makes it possible for them to carry out recognisance flights to new areas. Hence, repeated colonization of areas and colonization of new areas are expected more often by birds than by other animals. Here, we present data on the distribution of birds in the western Arctic, mainly the Canadian Arctic areas. We also include one site on Greenland, Nuuk. The observations were made during one summer, in 1999, and, therefore, we can exclude inter-year variations in weather as a factor influencing the results. The aim of the field-study was to collect data on the distribution and abundance of breeding birds and vagrant birds respectively. In the analysis, we will examine factors influencing the spatial distribution of different species and the occurrence of species associations.

Field-study

We collected data on the distribution and abundance of birds at 19 different sites in the western Arctic from Greenland to the western Canadian border. The easternmost site was at Nuuk, Greenland, and the most westerly site was in Ivvavik National Park, North Yukon, Canada. The site furthest to the north was on the Isachsen Peninsula, Ellef Ringnes Island, while Cap de Nouvelle-France on the Ungava Peninsula was the site furthest to the south. In all, 11 sites were on islands in the Canadian Arctic Archipelago and 6 sites (including King William Island) on the mainland, and yet another site was on Greenland. Tuktoyaktuk, here Tuk, on the mainland of Canada, was the site lying furthest away from the seashore. All other sites were fairly close to the seashore, within approximately 10 km or less. The expedition route, starting at the Ungava Peninsula on 1 July and ending at Cape Hooper, Baffin Island, on 1 September, enabled the collection of data on the distribution of birds during the breeding season and determination of the reproductive performance of many breeders up until the penultimate site at Devon Island. There several bird species such as peregrines, glaucous gulls and rough-legged buzzards were tending chicks during our stay. At the last site, Cape Hooper, it was late autumn with the winter migration of birds well in progress. Hence, we probably missed some bird species at this site, particularly those that regularly commence winter migration early. In addition, snow was falling during the survey which was carried out on foot, so the conditions at this site differed substantially from those at the other sites.

At each site, there were two groups, each consisting of three or four field-researchers, that surveyed the area for all birds. One of the groups concentrated their efforts on waders, catching and ringing individuals and determining reproductive performance. The other group surveyed the areas mainly for terrestrial species but included also observations of waders and ducks. With the exception of site 5 that was surveyed for 4 hours, each study area was censused in surveys carried out on foot over a period of 1–2 days, using binoculars and telescopes. We recorded the number of birds, the number of pairs of mated birds, the number of nests, the number of resident nonbreeding birds and the number of vagrant birds. The census routes were plotted on maps using topographic features as well as GPS-coordinates. Study areas were then plotted on maps, and the size of each study area determined using the grid-system on the maps. Moreover, aerial surveys were carried from a helicopter with two ornithologists on board. The helicopter flew at a speed of about 50 knots at approximately 50 m above the ground. Mammals and birds, the size of waders and larger, were determined according to species and the numbers counted.

In this preliminary report, the calculations do not include observations from Nuuk and Tuk (due to the locations of these sites), site 5, N. Bathurst, (due to the brevity of the stay), and Cape Hooper, Baffin Island, (due to the fact that winter migration ha d commenced).

Results and discussion

The number of species differed much between the sites e.g. the number of breeding species and potentially breeding species varied from 3 on Ellef Ringnes Island to 31 in the Ivvavik National Park, North Yukon. Six of the sites were on the mainland and eleven on islands, Nuuk and Tuk excluded. We recorded, on average, 20.8 ± 6. 7 species at the mainland sites with a range in species number of 15–31. On the islands, 14.4 ± 8.0 species were considered to be breeders and the range was 3–28 species. It would appear therefore that more species occurred at the coastal sites of the mainland than on the islands.

Moreover, we partitioned the whole area covered by the Tundra Northwest 1999 expedition in to four separate areas using a north-south divide from Ellef Ringnes Island to King William Island and an east-west divide following the Lancaster Sound, Viscount Melville’s Sound and McClure Strait. In the north-west area there was one site with 9 species while the north-east area comprised three sites (Ellef Ringnes Island excluded) with, on average, 12.3 ± 5.5 species. In the south-west area, which included Ivvavik National Park, the average number of species at six sites was 21.0 ± 7.6 species. Excluding data from King William Island, there were on average 18.0 ± 3.6 species at three sites in the south-east area. Hence, numbers of breeder species and potential breeders appear to differ in both north-south and east-west directions.

The bird data will be more rigorously analysed for differences in the distributions of bird groups and for associations between different species as well as for latitudinal and longitudinal variations.