Williamson's sapsucker (Sphyrapicus thyroideus) COSEWIC assessment and status report: chapter 5

• Previous page
• Table of contents
• Next page

Habitat

Habitat requirements

Williamson's Sapsuckers are found in coniferous mountain forests throughout their breeding range, usually at middle to high elevations in montane spruce-fir (Picea-Abies), Douglas-fir (Pseudotsuga menziesii), Lodgepole Pine (Pinus contorta), and Ponderosa Pine (Pinus ponderosa) forests (Dobbs et al. 1997). Mixed coniferous-deciduous forests are also used where Trembling Aspen (Populus tremuloides) can also be an important nesting tree (Crockett 1975; Smith 1982; Conway and Martin 1993, Loose and Anderson 1995). Hutto (1995) summarized 200 bird density studies in the northern Rocky Mountain region and found Williamson's Sapsucker to be widespread (detected on 10-36% of the studies in 7 of his 15 habitat classifications), and most common in dry conifer and post-burn habitats. However, in later studies, Hutto and Young (1999) indicated Williamson's Sapsuckers were absent from burns in their widespread surveys in western Montana and northern Idaho. Hutto and Young (1999) found them to be relatively uncommon in general in that area (not found on more than 2% of point counts in any habitat, and only at 0.3% of all the point counts), and most common in harvested forest cover types (partial cuts, patch cuts, seed-tree cuts) and in Cottonwood/aspen types, and less commonly in Mixed-conifer, Douglas-fir, and Wetlands.

Essential breeding habitat elements are:

• Trees suitable for nest cavity excavation, or existing tree cavities,
• Live coniferous trees for sap well creation, and
• Standing live and/or dead trees for gleaning ants (particularly carpenter ants, Camponotus sp.) from the bark surface to feed adults and nestlings in summer.

Stand structure and age

Stand structure does not appear to be as important as the individual habitat elements, i.e., these habitat elements can be incorporated into breeding territories in stands of quite differing structure, or in stands adjacent to each other that differ greatly in structure. Bull et al. (1986) described nest stand structure of Williamson's Sapsucker habitat in northeastern Oregon coniferous forests as basal area <34 m²/ha, <75% canopy closure, two or three canopy layers, and more than one dead tree per 0.1/ha. This definition appeared similar to the average “old-growth” Western Larch (Larix occidentalis) stand used for nesting in British Columbia (Gyug unpublished data, see later in this section). Stands were almost always multi-storied with a component of old (>200 yr) Western Larch, but most trees in the stand were much younger and smaller having originated after low or medium-intensity wildfires which the older trees survived.

In Colorado, Williamson's Sapsuckers favoured nest sites adjacent to open Ponderosa Pine forest (Crockett and Hadow 1975; Smith 1982) but actual nest stands were typically Trembling Aspen stands of 0.34 ha size with density of 772.4 trees/ha (range 182-1312 trees/ha; Crockett 1975). Conway and Martin (1993) found much lower aspen densities in nesting stands (mean 12.7/ha) probably because conifers made up the majority of the trees in the nesting stands. Crockett and Hadow (1975) and Smith (1982) found Williamson's Sapsuckers nesting in aspens to be foraging in nearby Ponderosa Pine stands, rather than in the aspen stands where they nested.

Stands used for nesting can vary from densely forested stands to very open stands with only occasional scattered trees. Detailed quantitative vegetation data for Williamson's Sapsucker nest stands in British Columbia has not been collected. Forest Cover mapping at 1:20,000 developed by the Ministry of Forests was used to describe the nest stands of 116 nests (Gyug unpublished data) in simple terms: 56.9% of the nest stands were Closed Forest (>30% tree canopy closure); 12.9% were at the edge of Closed Forest adjacent to an opening; 6.9% in Open Forest (10-30% tree canopy closure); 19.0% in Open areas (<10% tree canopy closure); either from logging or land clearing (8.6%), from wildfire (1%), or naturally open (9.5%). However, even though Open areas were used for nesting, almost all foraging trips observed from nests in Open areas were into nearby Forest stands. No nest was further than 140 m from a forested stand. The mean distance of nests from forested stands was 73 m (n = 21, SE = 8.4 m) for those nests in Open areas.

Stands containing old Western Larch appear to be disproportionately important to nesting Williamson's Sapsucker in the Okanagan-Greenwood population of thyroideus. Stand age and composition of either the nesting stand or the adjacent foraging stand were examined from Forest Cover mapping by the B.C. Ministry of Forests. Sixty-nine of 92 nests (75%) were in, or within 200 m of, stands where the primary canopy layer, or veteran tree layer, contained some Western Larch at least 170 years old (Gyug unpublished data). The proportion of Western Larch in the stands could be as little as 5%. A further 12.5% were in stands that contained some old (>170 years) or veteran Western Larch, but where this component had been missed in mapping, or was too uncommon to have been mapped. Stands containing old or veteran Western Larch trees accounted in total for 87.5% of the nesting/foraging stands in this population.

Williamson's Sapsucker nest and/or breeding pair density in large (>100 ha) census areas was proportional to the amount of old-growth forest, or old Western Larch trees, within the areas (Table 1). It appears that observed densities were the result of habitat quality, where the highest quality habitat is old-growth Western Larch forest because it contains old Western Larch for nesting and that contains carpenter ants for feeding on, and the stands are typically multi-storied with numerous smaller trees for sap well creation. Retaining the large Western Larch trees of that old forest, either in isolated trees in forested stands, or in patches, can maintain Williamson's Sapsuckers in an area, but only at a density in similar proportion to the amount of old-growth Western Larch forest or trees remaining.

In the Okanagan-Greenwood population in British Columbia, large areas without any suitable nesting Western Larch trees were found to be unsuitable habitat, even if sap well trees were available. There were no nesting Williamson's Sapsuckers in the Okanagan Falls Diameter Limit Cut census area (Table 1) even though the highest density of nesting pairs known in the province was immediately adjacent. This area had been diameter limit cut around 1971, where all conifers above a given diameter, which appeared to be around 25-cm DBH, were harvested. All cull trees and snags appeared to have been cut down as well. Even though there were Red-naped Sapsuckers nesting in this area in Trembling Aspen, no Williamson's Sapsuckers nested there.

¹ Gyug unpublished data
² Gyug and Bennett 1995;
³ Manning and Cooper 1996.

Stand size

Williamson's Sapsuckers in the Okanagan-Greenwood population in British Columbia used a variety of closed forest stand sizes, from as small as 0.5-ha to continuous forest (Gyug unpublished data). The minimum size of stand used for nesting was a Wildlife Tree Patch of 0.5 ha, although this was the exception rather than the rule. Most nests were either in much larger patches, adjacent (within 200 m) to larger patches, or in continuous forest.

At Wallace Creek, Williamson's Sapsucker populations were studied in 1995 and 1996 (Gyug and Bennett 1995; Manning and Cooper 1996). That area had contained some of the largest Western Larch in the entire Boundary Forest District (Randy Trerise pers. comm. 1994). Seed tree patches of a mean size of 3.6 ha (range 1.6-18.9 ha) had been retained throughout the upper drainage, covering about 10% of the gross area. Mean interpatch distance was 201 m. The area was logged in the late 1960’s by hand falling. Cull trees and snags, and small non-merchantable trees, were generally not cut. This was in contrast to later cutting in the area by mechanical harvesters where almost all standing trees and snags were felled unless reserving them was specifically part of the timber harvesting prescription. Williamson's Sapsuckers were found nesting in some of the patches, as well as in some of the Western Larch cull trees outside the patches, although most foraging appeared to be in patches of old-growth forest.

In the 1995 census of 24 patches at Wallace Creek, there were no nests or territory centres in patches of 1.6-2.0 ha (N=7; occupancy rate 0.00), 1 nest or territory centre in patches of 2.1-2.6 ha (N=8; occupancy rate 0.13) and 3 nests or territory centres in patches of 3.0-5.9 ha (N=9; occupancy rate 0.33) (Gyug unpublished data). There did appear to be a relationship between increasing patch size and occupancy by Williamson's Sapsucker, but no statistically significant relationship could be shown because of the high number of unoccupied patches. A similar relationship of Williamson's Sapsucker occupancy with increasing patch size of Ponderosa Pine forest was reported by Aney (1984) as cited in Dobbs et al. (1997). However, as at Wallace Creek, Williamson's Sapsuckers were probably too uncommon to show any statistical association with stand size since Hejl et al. (2002) did not pick out a minimum patch size that could be associated with occupancy by Williamson's Sapsuckers when citing Aney (1984).

Elevation range

Breeding range elevations used by Williamson's Sapsucker are relatively high at the southern end of the range, and decrease at the northern end of the range in British Columbia. Elevation range of breeding sites of thyroideus in British Columbia in the main Okanagan-Greenwood population was 860 to 1520 m (n = 89; Gyug unpublished data) with 88% of the nests between 1000 and 1390 m, 2% between 860 and 990 m, and 10 % between 1400 and 1520 m. Elevation range of breeding sites of thyroideus north and west of the Okanagan-Greenwood population (n = 27) was 700 to 1250 m (Gyug unpublished data; BCNRS; B.C. CDC; WLAP). The majority (81%) were between 800 and 1100 m. Elevation of 10 nests not included in the above summary near Merritt and Hat Creek (Cooper 1995) ranged from 850 - 1100 m.

The elevation range of breeding sites occupied by nataliae in British Columbia is not known. There was insufficient precision in the locations of historical nesting records to estimate an accurate elevation range. The only known nest site occupied in 2004 was about 1100 m elevation.

Biogeoclimatic ecosystem classification

The entire province of British Columbia has been classified and mapped into biogeoclimatic zones, subzones, and variants as part of the Biogeoclimatic Ecosystem Classification (BEC, e.g. Lloyd et al. 1990). Williamson's Sapsucker nests in the Okanagan-Greenwood population of thyroideus occurred primarily (70%) in the Kettle Dry Mild Interior Douglas-fir variant (IDFdm1, Table 2). In other populations, the nests, or likely breeding locations, were primarily (85%) found in the Okanagan Very Dry Hot Interior Douglas-fir variant (IDFxh1), its grassland phase (IDFxh1a), and the Thompson Very Dry Hot Interior Douglas-fir variant (IDFxh3).

¹ Based on Version 5 BEC Classification (2003), and on draft Version 5 BEC Classification for Kamloops Forest Region provided by Dennis Lloyd (Ministry of Forests, Kamloops), see Lloyd et al. (1990) for description of BEC Classification variants and phases.
² Gyug unpublished nest site data.
³ Nest sites (N = 20), drumming males where no return was made to search for nests (N = 3) from Gyug unpublished data; from BCNRS (N = 6); WLAP (N = 2); and CDC (N = 2) where location was known with enough precision to determine BEC Classification of site.

Breaking down the mapped “breeding range” (a subset of the Extent of Occurrence eliminating large areas of apparently unsuitable or unused habitats) and the Areas of Occupancy by BEC Units showed that it is not a random selection of BEC units from within the “breeding range” (Table 3). The IDFdm1, the IDFxh1 and the MSdm1 represent a total of only 15% of the “breeding range” but account for 92% of the Area of Occupancy of the Okanagan-Greenwood population. The IDFdm1 and the MSdm1 are distinguished from other similar BEC units by the presence of Western Larch. The Area of Occupancy of the other populations include PPxh1 and IDFdk1 in approximate proportion to their overall abundance in the “breeding range” but a disproportionately larger amount of IDfxh1, IDFxh3 and IDFxh3a (68%) compared to general abundance in the “breeding range” (21%). The precise locations of the nataliae nest records were not known, but were all either in Kootenay Dry Mild Interior Douglas-fir variant (IDFdm2) or Dry Cool Montane Spruce (MSdk).

¹ Breeding Range, the shaded area of Figure 2, is a subset of the Extent of Occurrence, eliminating large areas of apparently unsuitable or unused habitat.
² BEC Units listed in the table from lowest to highest elevations.

Range limitation by habitat requirements

The only dense populations of Williamson's Sapsucker in British Columbia appear to be associated with forest stands that have a component of old veteran Western Larch in them (also see Biology, Nest Trees section). However, small and/or low density populations do occur outside of this area in association with Trembling Aspen and/or Ponderosa Pine stands. It is not known exactly why there might be so few Williamson's Sapsuckers in these other areas, or what may be preventing further expansion of their range into what appears to be an abundance of aspen habitats in southern British Columbia when the same species uses aspen habitat extensively in the southern Rocky Mountains.

If the only habitat requirements of the species were for Trembling Aspen nesting trees, Douglas-fir sap trees, and old Douglas-fir trees or snags with an abundance of carpenter ants for feeding nestlings, then there are many areas in the southern interior of British Columbia that fit this description. However, these appear to be unused by Williamson's Sapsucker. Pileated Woodpeckers have similar requirements for carpenter ants (Bull and Jackson 1995) and large nesting trees (of a variety of species including Trembling Aspen) but are found in a far wider range of habitats than Williamson's Sapsucker, indicating that if these were the only habitat requirements, then Williamson's Sapsucker would probably be much more widespread and/or abundant. The critical factor may be for old-growth Ponderosa Pine stands, which are present in some occupied locations of the Princeton and Merritt populations, but are currently relatively rare throughout the southern interior of British Columbia because these were logged over the past 100 years.

Habitat trends

The primary population centre of thyroideus in British Columbia is in the Okanagan-Greenwood population where 85% of the estimated Total Population occurs (see Population Sizes and Trends section below). Loss of old-growth Western Larch stands to logging and land clearing is the primary habitat threat, with extensively logged areas and intensively logged stands unsuitable for breeding. Habitat loss from timber harvesting and land clearing in this range is continuing. Based on the B. C. Broad Ecosystem Inventory (BEI) and adjustments for recent forest harvest and fires, the report writer estimated that the Williamson’s Sapsucker Area of Occupancy contains about 89 km² (15.0%) of older forests, and that over the past 10-14 years, there has been a loss of 23% of the older forests in this Area of Occupancy.

Land clearing on private land at Anarchist Mountain has contributed to the loss of local breeding sites in the Okanagan-Greenwood population where they were formerly common (see Population Trends below). In the 1990’s the private land property taxation methods were changed in British Columbia. The standing timber on the land now forms part of the property value to be assessed, and annual property taxes on rural land are based on this new assessment (B.C. Assessment 2004). There is now very little incentive to keep standing timber on private land because it is being taxed without providing any revenue unless cut down, and is not taxed as part of the property value once it has been cut down.

In the Princeton population, the majority of trees in the old-growth (>200 yr old) Ponderosa Pine nesting stand at August Lake were harvested in 2003 to combat Mountain Pine Beetle (Dendroctonus ponderosae) infestations. While a pair was in the area on 25 May, 2004, they were not present on 14 June, and nesting was probably not successful (Gyug unpublished data). This stand had been noted as the most reliable place to find nesting Williamson's Sapsuckers in the Princeton area for over 15 years (BCNRS, B.C. CDC, Jerry Herzig pers. comm. 2000, Dick Cannings pers. comm. 2004, Gyug unpublished data).

Also in the Princeton population, nesting had been recorded in 1984 (BCNRS) at Whipsaw Creek. This old-growth Ponderosa Pine stand on private land south of Princeton was logged in the spring of 2004. Every large Ponderosa Pine was removed from the site, and the habitat rendered completely unsuitable for Williamson's Sapsucker. An adjacent Ecological Reserve is too small (29 ha), has few suitable old-growth Ponderosa Pines and would probably not be suitable habitat for Williamson's Sapsucker on its own.

Loss of habitat to timber harvesting is ongoing. Timber harvesting on Crown lands from 1997-2003 has resulted in the loss of local sites in lower Ingram Creek north of Midway in the Arrow-Boundary Forest District that appear to have been occupied in the past (Gyug unpublished data from 1996; WLAP unpublished data) but are now unoccupied (Gyug unpublished data from 2003 and 2004). In other areas, the 2004 field surveys (Gyug unpublished data) found the area around 6 active nests to have been surveyed out for future timber harvesting, 3 to have been logged within the past 2 years but with enough residual standing timber and nest trees remaining to keep Williamson's Sapsuckers nesting (at least on the short-term), and 2 of the 4 areas occupied by apparently unmated males to be surveyed out for logging.

Timber harvesting has been proposed, and the boundaries have been surveyed, for about 30% of the forested area of the Okanagan Falls Old-Growth Western Larch census area which is approximately 230 ha in size (see Table 1). This site has the highest density of Williamson's Sapsuckers known in Canada and in its entire range, and is already bounded by unsuitable logged habitat. The site (formerly known as Schoonover Mountain, now most commonly referred to on BCINTBIRD birder’s e-mail group as Venner Meadows Road) has been known as one of the most reliable places to find Williamson's Sapsuckers in Canada for 90 years. Based on personal knowledge of the Okanagan-Greenwood Area of Occupancy and other areas throughout the Okanagan surveyed for this and other species and projects from 1990 to 2004, review of aerial photography of the entire area, review of the digital forest cover mapping for the Okanagan-Shuswap Forest District, Boundary TSA and Merritt TSA, and the Terrestrial Ecosystem Mapping of Tree Farm Licence 15, there are no other old growth Western Larch patches of this size (>200 ha), age (>200 yr) and apparent habitat quality for Williamson's Sapsucker in the Area of Occupancy of the thyroideus subspecies in British Columbia. The site would appear to be irreplaceable.

Timber harvesting rates can be estimated at about 1% of the forest land base per year using a very simplistic model based on clearcutting, and on an approximate 100-year rotation age between harvesting entries in any given location. Actual projected timber harvesting rates for the last year (2002) in which they were publicly available on the British Columbia Ministry of Forests website (they are no longer publicly posted as GIS files that can be overlaid over other GIS habitat layers), were 1.26% per year for the portion of the thyroideus Williamson's Sapsucker Area of Occupancy in the Okanagan-Shuswap Forest District. This was calculated based on the five-year planning timeframe of 2002-2007 in the Area of Occupancy (233.8 km²) where 12.97 km² or 6.3% of the Crown Land portion (205.7 km²) of this Area of Occupancy was proposed for timber harvesting in that time period. Most of this area proposed for timber harvest (10.27 km² or 79%) was in forests with an older Western Larch component in general areas known to be occupied by breeding Williamson's Sapsuckers.

The old trees that are the primary nesting sites and ant foraging sites may take hundreds of years to reestablish, given that the veteran Western Larch nesting trees are typically 200 to 600 years old. Of 18 nest trees found in 1996, 1997 or 1998, and followed until 2000, 5 were used for more than one year of nesting with one used for 4 years in a row (Gyug unpublished data). Three of these trees fell over in that period--a Black Cottonwood (Populus balsamifera subsp. trichocarpa), a birch (Betula sp.) and a Ponderosa Pine--but none of the 11 Western Larches fell over. In general, Western Larch nest trees may be longer lived, and therefore more stable and predictable, nesting substrates than other tree species in this area.

There are also threats to individual trees, as well as to whole stands. One veteran Western Larch nesting tree used from 1997 to 1999 was felled by firewood cutters in 2003. The tree had extensive heart rot, a dead and broken off top, and extensive burn and perhaps lightning scars on its trunk, but was still alive. About half of the tree was abandoned on site because it was too rotten to cut with a chainsaw.

While habitat is being lost, no new habitat is being created. Most timber harvesting in the area does not leave enough residual standing timber to satisfy Williamson's Sapsucker habitat requirements. If there were no adjacent old-growth Western Larch/Douglas-fir stands for foraging and/or nesting, then the habitats were generally found to be unoccupied (Gyug unpublished data). In the 1980’s and 1990’s, the Ministry of Environment (now the British Columbia Ministry of Water, Land and Air Protection) negotiated extensive leaves of mature Western Larch in clearcuts at densities up to 60-100/ha, but more typically 5-10/ha. While initially forest licensees had been leaving these as seed trees for natural regeneration of clearcuts, the regeneration was too unpredictable and has given way to the widespread planting of seedlings on all cutover sites. The value of these mature seed trees was recognized by wildlife managers in the 1980’s and 1990’s, and these trees were being reserved from harvest, even though they no longer were required for conifer regeneration.

However, in the past few years the reserve of mature trees from harvest in clearcuts appears to have changed with the advent of the British Columbia Forest Practices Code (now replaced by the Forest and Range Practices Act) and with the imposition and enforcement of Worker’s Compensation Board (WCB) safety rules. The Wildlife Tree Guidelines developed under the Forest and Range Practices Act have been developed separately for each Forest District to recognize the unique situations in each district. However, these are generally area-based and only recognize patches of trees as meeting requirements. Areas that would have had Western Larch or Douglas-fir reserved only a few years ago in partial imitation of the ground-fire disturbance regimes which would have naturally left many of these older trees standing are now generally cutover completely, with no standing residual timber. Salvage harvesting of areas burned in wildfires in 2003 are coming under the same system, with many large Western Larches surviving the fire, but being salvage logged along with the standing dead trees.

The WCB safety rules apply to any stands where people or machinery are working, and generally do not permit the retention of trees with obvious defects, which would formerly have been cull trees and often left standing. Large veteran Western Larch nesting trees--most of which have obvious defects such as fire scars and dead or broken tops--as well as the trees that may have become the next generation of nesting trees, are a quickly disappearing component of the timber harvesting landscape in British Columbia.

Habitat trends in the nataliae population have not been quantified in the 1996-2004 studies which only examined the thyroideus populations. However, Johnstone (1949) identified logging in 1939 as the agent which destroyed the habitat at the only nesting site known to be occupied in 1938. Only 95.6 km² (3.4%) of the nataliae breeding range was estimated to be in forests older than 140 years as of the early 1990’s based on Broad Ecosystem Inventory. This is much lower than the estimated 19.5% suitable habitat in the Okanagan-Greenwood population for the same time period.

Habitat protection/ownership

Only one currently occupied nest site of Williamson's Sapsucker in British Columbia is in a park. The nest near Kimberley in 2004 was in a park of leased Crown land managed as a park by the City of Kimberley (Kimberley Nature Park Society 2004). Two other nesting sites, one at Lightning Lakes in E.C. Manning Provincial Park (BCNRS), and one at Cathedral Provincial Park (Campbell et al. 1990) were not occupied in 2004, and seem not to have been occupied for a number of years. Some appear to be on private ranch land at Hat Creek (although the exact location of the nests was not available), and at Merritt and Princeton, and along Highway 3 at Anarchist Mountain (Gyug unpublished data). All other known sites are in Crown-owned forest.

No breeding sites are known from any First Nations Reserve lands. The only First Nations Reserve Lands within Williamson's Sapsucker Area of Occupancy were 11 km² in the Okanagan-Greenwood population east of Oliver. These particular areas have not been inventoried for Williamson's Sapsucker, but certainly may contain Williamson's Sapsucker based on nests in similar habitat <200 m from these lands.

Planning is currently underway for Old-Growth Management Areas (OGMA) as reserves from timber harvesting under the Biodiversity Guidelines of the Forest and Range Practices Act. Sixty-five of the currently planned 1421 OGMAs (draft 3, October 2003, coverage provided by British Columbia, Ministry of Sustainable Resource Management) in the Okanagan-Shuswap Forest District intersect the Williamson's Sapsucker Area of Occupancy in the Okanagan-Greenwood population. However, only one of these OGMAs contains a known Williamson's Sapsucker nest site. There were also two other observations of Williamson's Sapsuckers in other draft OGMAs: one at Anarchist Mountain, and one at Micah Creek. Only four of the 50 known nest sites in the Arrow-Boundary Forest District were within draft OGMAs (May 2004 version available at http://tfic1. timberline.ca/ kbhlpo/index/boundary.html). In general, the currently planned OGMAs will do little to protect Williamson's Sapsucker populations