Skip booklet index and go to page content

COSEWIC Assessment and Update Status on the Northern Goshawk (2000)

Population Size and Trends

Inventory Results

Goshawks are a rare to uncommon forest raptor throughout forested parts of British Columbia; until recently, provincial records of nesting were relatively few (Campbell et al. 1990). During the last few years, inventories of nesting goshawks have occurred in several regions of British Columbia in response to the high conservation priority placed on this species. Surveys in coastal British Columbia located 41 Queen Charlotte Goshawk nests on Vancouver Island during 1991-1999: 1 nest-1991, 4 nests- 1994, 6 nests-1995, 8 nests-1996, 9 nests-1997, 7 nests-1998, 6 nests-1999 (McClaren 1999; D. Doyle pers. comm.). On the Queen Charlotte Islands, 5 active nests have been reported during 1995-1999: 1 nest-1995, 2 nests-1996, 0 nests-1997, 1 nest-1998, 1 nest-1999 (Chytyk and Dhanwant 1999; Chytyk et al. 1999).

Population Estimates

A habitat capability model developed for the Queen Charlotte Goshawk, estimated 1,700 pairs in coastal British Columbia and 800 pairs in southeast Alaska (Crocker-Bedford 1990b). The British Columbia estimate included the coastal mainland even though it is uncertain if the Queen Charlotte Goshawk occurs there. More recent refinements to the model reduced the estimate for southeast Alaska to 100-200 pairs (Crocker-Bedford 1994). Although a new estimate for British Columbia was not made, Crocker-Bedford (1994) suggested the British Columbia population was substantially less than 1,700 pairs. This conclusion seems reasonable given the results of recent surveys on Vancouver Island (McClaren 1997, 1998, 1999) and the Queen Charlotte Islands (Chytyk et al. 1998; Chytyk and Dhanwant 1999). However, the accuracy of these estimates are questioned by other goshawk experts. Even so, results from the above-mentioned surveys suggest that densities, although uncalculated as yet, on Vancouver Island are possibly much higher than on the Queen Charlotte Islands.

If we use the maximum of Crocker-Bedford’s (1994) revised population estimate for southeast Alaska (200 pairs), and apply the same reduction (25% of his 1990 estimate) to his original British Columbia estimate (1,700 pairs), then there would be an estimated 425 pairs of Queen Charlotte Goshawks in coastal British Columbia. 

The Queen Charlotte Goshawk is known to nest on large coastal islands (West Cracroft, Quadra and E. Thurlow islands (E. McClaren pers. comm.) and perhaps the north and central coastal mainland of British Columbia (Campbell et al. 1990; Crocker-Bedford 1994). The coastal mainland is comprised of many rugged small islands and islets and steep glacial fjords and high alpine areas. These terrain types offer few areas of preferred Queen Charlotte Goshawk nesting habitat, i.e., lower elevation moderate slopes (McClaren 1999; Chytyk and Dhanwant 1999). It is thought that the drier biogeoclimatic zones on the east side of Vancouver Island contain the best Queen Charlotte Goshawk nesting habitat in British Columbia (D. Doyle pers. comm.). The coastal mainland is somewhat wetter than eastern Vancouver Island and, possibly, does not offer as good quality nesting habitat as is found on Vancouver Island.

We estimate that about 10% of Vancouver Island, mainly in the northeast, has been inventoried for nesting Queen Charlotte Goshawks (D. Doyle pers. comm.). Forty-one Queen Charlotte Goshawk nests have been reported by McClaren (1999) during the last 5 years. If we assume that those 41 nests have been found in about 10% of the landscape, that the area surveyed is relatively high quality goshawk, and the remainder of unsurveyed habitat is somewhat lower quality, then a reasonable population estimate for Vancouver Island may be about 300 pairs.

We suggest that a conservative estimate of the relative density of Queen Charlotte Goshawks on the Queen Charlotte Islands is about 50% of that found on Vancouver Island, for the following reasons:

  1. Generally, it is thought that goshawk territories increase in size the higher the latitude (Squires and Reynolds 1997). The Queen Charlotte Islands is almost 4o latitude further north of Vancouver Island and, consequently, territory size on the Queen Charlotte Islands is likely slightly larger than on Vancouver Island. On Vancouver Island, the closest inter-territory distance is 3.2 km, while a cluster of 5 territories averaged approximately 6 km apart (McClaren 1999). Using the average of 6 km inter-territory distance as the diameter of a circular territory, a rough estimate of territory size for Vancouver Island would be about 2800 ha. Territory sizes in southeast Alaska and, based on our impressions, probably the Queen Charlotte Islands, appear to be considerably larger than those suggested by data for Vancouver Island.

  2. The Queen Charlotte Islands contain proportionately less potential nesting habitat than Vancouver Island. The northeast portion of Graham Island, the northern of the two large islands on the Queen Charlotte archipelago, is a large area of stunted lodgepole pine forest and open bogs, generally considered low quality nesting habitat (Johnsgard 1990; Squires and Reynolds 1997). The southern portion of Moresby Island, the southern of the two large islands, is mostly rugged, small islands and islets, that do not contain many moderate slopes or flatter valley bottoms, preferred nesting habitat of Queen Charlotte Goshawks (McClaren 1999; Chytyk and Dhanwant 1999). Relatively, Vancouver Island does not have many stunted forests, islands or islets, however, it does have a greater proportion of alpine and subalpine areas. As well, the majority of second growth found on the Queen Charlotte Islands is relatively young and generally not suitable for nesting habitat (Chytyk and Dhanwant 1997). Large-scale logging did not start to impact the landscape significantly until the mid to late 1960s on the Queen Charlotte Islands and, consequently, there are few large areas of second growth >50 years old. This contrasts with Vancouver Island where the rate of commercial logging accelerated in the 1940s and, as a result, there is significantly more areas of second-growth forest old enough to provide nesting habitat.

  3. There is a smaller prey base on the Queen Charlotte Islands than on Vancouver Island. Provincial records show that several key prey species of grouse, woodpecker and medium-sized songbird are absent from the Queen Charlotte Islands, but are present on Vancouver Island (Campbell et al. 1990, 1997). The absence of these species on the Queen Charlotte Islands may have a more dramatic effect on Queen Charlotte Goshawk populations during winter months (Chytyk and Dhanwant 1999) when starvation is a major cause of death, as noted on Vancouver Island (McClaren 1999).

  4. Population inventories conducted on Vancouver Island and the Queen Charlotte Islands provide some evidence for lower population densities on the Queen Charlotte Islands. Inventories have been conducted on Vancouver Island between 1994-1999 and on the Queen Charlotte Islands between 1995-1998. During the two inventory periods, 41 nests were found on Vancouver Island while 5 nests (approximately 12% of 41) were found on the Queen Charlotte Islands. Because survey efforts were greater on Vancouver Island than on the Queen Charlotte Islands, but not sufficiently greater to account for the much higher success in nest detection, we suggest densities on the Queen Charlotte Islands are lower rather than higher.

The Queen Charlotte Islands has an area of 9,596 km2 and is roughly 31% the size of Vancouver Island at 31,284 km2. Using the 31% geographical area proportion and the conservative estimate of the population density on the Queen Charlotte Islands being half that of Vancouver Island, then, we suggest that the Queen Charlotte Islands has an estimated population of approximately 50 pair of Queen Charlotte Goshawks. These estimates suggest that about 14% (41 of 300 estimated pairs) on Vancouver Island, and 10% (5 of 50 estimated pairs) on the Queen Charlotte Islands of the population has been documented by surveys. These percentages seem reasonable given the areas surveyed, the results of those surveys, and the area remaining to be surveyed.

The status of goshawks breeding on the coastal mainland is uncertain. They may be Queen Charlotte Goshawks, as assumed by Crocker-Bedford (1990, 1994), A.g. atricapillus, or a mixture of both. However, if we ignore the possible population on the mainland coast and other coastal islands, we suggest that a conservative population estimate of Queen Charlotte Goshawks in British Columbia is 350 pairs. If we include the mainland coast and other coastal islands, then we estimate >425 pairs.

Population Trends

Population trends are unknown in British Columbia and, apparently, anywhere else in western North America. Some studies which report declines have apparently not applied sufficiently rigorous statistical methods to determine trends (Kennedy 1997). However, Crocker-Bedford (1998) and Smallwood (1998) point out that it may be nearly impossible (due to practical factors such as costs) to obtain a data set for this raptor species with sufficient rigour to prove any trend statistically. For the Queen Charlotte Goshawk, Crocker-Bedford (1990b) estimated that habitat for 1150 pairs has been converted to early seral forest in southeast Alaska and coastal British Columbia due to logging of old-growth forests. Even if Crocker-Bedford’s predictions were inaccurate (Kennedy 1997, 1998), he suggested (Crocker-Bedford 1998), and we agree, the trend seems to be inevitably downward for Queen Charlotte Goshawks in British Columbia because of the extent of logging of old-growth and mature second-growth forests on the coast.