Skip booklet index and go to page content

COSEWIC Assessment and Status Report on the Red Crossbill (Percna) in Canada

Population Sizes and Trends

Historical Trends

Although no quantitative information exists, Red Crossbills were formerly considered abundant in Newfoundland. In the mid-19th century, Reeks (1869) described them as common early nesters on the west coast of Newfoundland in the northern portion of what is now Gros Morne National Park. 

Around the turn of the 20th century, many songbird specimens were collected in western, central and eastern Newfoundland (Montevecchi and Tuck 1987). Interestingly, among the specimens collected in Newfoundland in the 1890s and early 1900s, Red Crossbills outnumber White-winged Crossbills by more than a 2:1 ratio (21:9). If collectors did not favour Red Crossbills over White-winged Crossbills (they may have as they targeted distinctive sub-species though they seemed to collect everything) and if Red Crossbills were not more vulnerable to collectors, then it is possible that Red Crossbills could have been more common in Newfoundland than White-winged Crossbills around the turn of the century (WAM unpublished files). 

Noble (1919) wrote in 1915 “flocks of Crossbills were observed flying over the Humber on many different evenings.” Rooke (1935) described White-winged Crossbills as “commoner” than Red Crossbills, and Peters and Burleigh (1951) considered the Red Crossbill in Newfoundland as fairly common locally in summer, common in the Codroy Valley in September, but uncommon in winter, with erratic and local distribution.

Red Crossbills were considered to be “regular” and “similar” in abundance to White-winged Crossbills in western Newfoundland during the late 1950s/early 1960s (H.Deichmann unpublished records). Erskine (1977) describes Red Crossbills as “common” in spruces in Newfoundland in 1968. In the late 1960s and 1970s, nests and fledglings of Red Crossbills were observed in the St. John’s area and elsewhere in eastern Newfoundland; no nests have been observed since 1977 (WAM unpublished files) but juveniles have occasionally been observed in this area in recent years. 

By 1975, White-winged Crossbills were considered to be more common than Red Crossbills in Gros Morne National Park in western Newfoundland (Lamberton 1976). The 1982 and 1989 Newfoundland bird checklists (Maunder and Montevecchi 1982; Mactavish et al. 1989) list both White-winged and Red Crossbills as uncommon (likely to be found monthly in appropriate season/habitat; may be locally common), breeder (known to breed), resident (non-migratory, or maintains a significant year-round population). The 1999 and 2003 checklists (Mactavish et al. 1999, 2003) classify White-winged Crossbills as common (likely to be found daily in appropriate season/habitat), irruptive (irruptive species, much more abundant in some years than in others), breeder, resident; while Red Crossbills are designated very uncommon (likely to be found annually in appropriate season/habitat; may be locally uncommon), breeder, resident.

 

Recent Trends

It is difficult to estimate the populations of songbirds. It can be particularly difficult to determine the populations and population fluctuations of irruptive species like crossbills, because they often move in large flocks of thousands of birds over continental North America (Bock and Lepthien 1976). Populations can fluctuate with periodic cone failures and abundances (Koenig and Knops 2001; Wren 2001). As a result, it is essential to analyze long-term trends over large spatial scales in irruptive species like crossbills so that multi-year patterns in population trends can be discerned.

Christmas Bird Counts

The longest and most consistent source of information on population trends for Red Crossbills in Newfoundland comes from Christmas Bird Counts (CBCs). CBC methodology involves day-long counts performed annually within a 2-week period of Christmas in hundreds of discrete 24.1 km diameter circles across the continent. Numbers of all birds encountered in each circle are recorded. These counts have been run at 11 sites on insular Newfoundland at some point during the past 5 years (Figure 5), but additional counts in different locations on the island have been performed since CBCs started in Newfoundland in the 1960s. The total number of Red Crossbills counted on all Newfoundland CBCs is presented in figure 6. These data are displayed both in raw number form, and standardized by the number of party-hours in the field to control for variable observer effort (Bock and Root 1981). [Note that each CBC year spans two calendar years, thus for example, the CBC year 1996 represents the count performed during a two-week period in December 1995/January 1996. This method of labelling is used throughout this document.] 

While the irruptive ecology of Red Crossbills makes them unlikely to turn up on CBCs with annual regularity, it is apparent from Figure 6 that they have declined greatly, and over the past decade they have not been counted in numbers close to those of high-abundance years in the past. By fitting exponentials to the curve using non-linear least squares regression, an exponent of decay was used to calculate the annual rate of decline of Red Crossbills from this data series. The analysis indicates a 99% decline of number of Red Crossbills per party-hour over the period from 1968 – 2002 (calculated based on the decay exponent -0.1369 from the regression), and a 10-year rate of decline of 75%. This trend analysis changes the count per party-hour from the outlier year 1980 to the mean of its neighbouring values; with that year included in the analysis, the total time period decline is 98% and the 10-year decline is 67% (decay exponent -0.1122).

Figure 5:  Locations of Christmas Bird Count (CBC) Circles on Insular Newfoundland Active 1997 - 2002

Figure 5.  Locations of Christmas Bird Count (CBC) circles on insular Newfoundland active 1997 - 2002. Newfoundland coastline mapã2000, Government of Canada with permission from Natural Resources Canada (2000).

Newfoundland coastline map © 2000, Government of Canada with permission from Natural Resources Canada (2000). CBC count circle geographic information obtained from National Audubon Society (2002).

The longest records of CBCs in Newfoundland come from St. John’s on the east coast and from Terra Nova National Park on the northeast coast of the island. The CBCs from Terra Nova National Park show a relatively consistent decline in Red Crossbills from the late 1960s through 2003 (Figure 7). Counts were significantly higher during 1968 through 1973 than from 1974 through 1988 (Mann-Whitney Test, U = 3.5, n = 21, P < 0.002; compare 1968-73 with 1974-88). White-winged Crossbills have also exhibited considerable fluctuations over this same time period but their highest count was obtained in 1996 (Figure 7), an extraordinary cone year in Newfoundland, during which White-winged Crossbills were sighted in large flocks from all parts of the island. There was no increase of sightings of Red Crossbills in Newfoundland during the winter of 1995-96 (WAM unpublished files). In other parts of their range, Red Crossbills and White-winged Crossbills typically irrupt with synchrony (Bock and Lepthien 1976), so it may not be unreasonable to expect high numbers of Red Crossbills in Newfoundland to be normally found during years with high White-winged Crossbill numbers. A similar declining trend for Red Crossbills is demonstrated in the CBC data from the St. John’s count. Consistent with the Terra Nova data, White-winged Crossbill abundance has been erratic and some recent CBC counts have had quite high numbers of White-winged Crossbills per party hour, but no Red Crossbills.

Figure 6:  Total Number of Red Crossbills Per Party Hour Observed on All Newfoundland CBCs, 1967 – 2002

Figure 6.  Total number of Red Crossbills per party hour observed on all Newfoundland Christmas Bird Counts, 1967 – 2002 (data obtained from National Audubon Society 2002).

Data obtained from National Audubon Society 2002.

Because there is evidence that Red Crossbill percna subspecies may be found in mainland areas such as Anticosti Island and Nova Scotia (although with infrequent regularity and unknown breeding status), it may be useful to include Red Crossbill counts from these areas when considering percna trends. CBC data are available for Nova Scotia, but not Anticosti. While CBC data provide no information on subspecies counted in the field, it is possible that some Red Crossbills counted on Nova Scotia CBCs are the percna subspecies (although CBC data from Nova Scotia will overestimate numbers of percna because the majority of birds recorded are likely other subspecies). CBC data from Newfoundland and Nova Scotia combined indicates a declining trend in Red Crossbills (Figure 8). Regression analysis of log-transformed count per party hour indicates a 90% decline over the period from 1968 – 2002 (slope of line = -0.0281, R2 = 0.4064), and a 10-year rate of decline of 49%. This trend analysis changes the count per party-hour from the outlier year 1980 to the mean of its neighbouring values; with that year included in the analysis, the total time period decline is 92% (slope of line = -0.0307, R2 = 0.3642) and the 10-year decline is 51%. Because CBC data indicate that all Red Crossbills censused in Newfoundland and Nova Scotia have greatly declined, it implies that the unknown portion of Nova Scotia Red Crossbills of the subspecies percna which are counted with other Red Crossbills have likely also declined.

Figure 7:  Number of Red and White-winged Crossbills Per Party Hour on St. John’s (1967 – 2003) and Terra Nova National Park (1969 – 2003) CBCs


Figure 7.  Number of Red and White-winged Crossbills per party hour on St. John’s (1967 – 2003) and Terra Nova National Park (1969 – 2003) Christmas Bird Counts (data obtained from National Audubon Society 2002).

Data obtained from National Audubon Society 2002.

Figure 8:  Total Number of Red Crossbills Per Party Hour Recorded on All Newfoundland and Nova Scotia CBCs, Standardized by Total Party Hours for Both Provinces, 1969 – 2003

Figure 8.  Total number of Red Crossbills per party hour recorded on all Newfoundland and Nova Scotia Christmas Bird Counts, standardized by total party hours for both provinces, 1969 – 2003. 

Note: high count at 1980 due to one single large flock of birds at Cape St. Mary’s, Newfoundland. Data obtained from National Audubon Society (2002).

Although CBCs are the longest running census providing information on the Red Crossbill, there are difficulties associated with using sparse (i.e., one day a year and restricted geographic extent) census data collected by variable numbers of observers of variable capability to analyze long-term trends in an irruptive species like the Red Crossbill (Dunn and Sauer 1997). The CBC data provide a snapshot of early winter bird occurrences in restricted geographic areas (the 24.1 km diameter prescribed count circles). The nomadic nature of Red Crossbills means that their abundance over an entire winter season may not be accurately sampled by CBC methodology (crossbills could be counted in a large flock which does not represent their abundance over a larger geographic area, or crossbills could fail to turn up on a census when they are relatively abundant in nearby areas outside the count circle). However, it might also be expected that large foraging flocks of conspicuous, irruptive species like crossbills might be detected more consistently than other rarer and less conspicuous species. It is important to remember the snapshot nature of the data when assessing crossbill abundances from CBCs, and to consider the data primarily as a source of information on gross trends.

Breeding Bird Surveys

Breeding Bird Surveys (BBS) have been conducted annually in Newfoundland since 1973. The BBS is a standardized count regime that is run in June and early July in Newfoundland. Skilled observers drive a roadside survey route of 39.4 km, stopping every 0.8 km to record all birds heard/seen within a 0.40 km radius (USGS Patuxent Wildlife Research Center 2003). There are 25 routes with random start points in every degree block with sufficient roadway on insular Newfoundland. Some of these routes occur in primarily forested areas of the province: for example, 2 routes are found in the Western Newfoundland ecoregion, 6 in the Central Newfoundland ecoregion, and 3 in the Northern Peninsula Forest ecoregion.

Analysis of Red Crossbills counted on all Newfoundland BBS routes, 1966 – 2000, indicates a relative abundance of 0.4 birds/route, and a declining trend of 14.4% per year (Sauer et al. 2001). However, this trend is not significantly different from a trend of 0% change per year, as the analysis is based on only four routes, and the results are highly variable (a very large 95% confidence interval). Thus, records from Newfoundland BBSs are not useful in analysing long-term population trend changes of Red Crossbills. This may be partly attributable to the irregular appearance of crossbills (both Red and White-winged) on surveys: that they are not reliably present in a given location from year to year complicates this type of population trend analysis (Link and Sauer 1998).

Newfoundland BBSs can, however, be used as an annual measure of presence/absence and of distributional occurrences/habitat associations of Red Crossbills. Red Crossbills were reported in eight consecutive years of BBSs (1979 – 1986) and since then have been reported only twice; however, coverage and temporal sampling of routes has been poor (USGS Patuxent Wildlife Research Center 2003). 

Other Accounts

The Newfoundland Forest Service conducts regular summer/early fall wildlife surveys in a variety of forest habitats across the island. In 1,426 surveys over the period from 1996 to 2002, Red Crossbills have only been observed twice (one bird on August 16, 1998 in a Balsam Fir stand in central Newfoundland, and one bird on June 25, 2000 in a second Balsam Fir stand in central Newfoundland; C. Cohlmeyer, Newfoundland Forest Service data files, 2003).

Other recent records of Red Crossbill sightings also exist. For example, birdwatchers from across the province report bird sightings to the newsgroup nf.birds (2003), where Red Crossbills are reported irregularly but not infrequently (Figure 9). Of note are the sightings of Red Crossbills from April/May 2002, when a flock of at least 12 birds was repeatedly observed foraging in non-native pines (P. nigra) planted around the campus of Memorial University of Newfoundland. This may partly explain the increase in reporting of Red Crossbills to this newsgroup (as the Newfoundland birdwatching community is centred in St. John’s); the increasing trend could also be partly explained by the increased users of the newsgroup in more recent months. Red Crossbills are also occasionally reported on the French islands of Saint Pierre et Miquelon (R. Etcheberry, pers. comm., 2003). 

Figure 9: Red Crossbills Observed in Newfoundland Per Month, 1998 – 2002

Figure 9. Red Crossbills observed in Newfoundland per month, 1998 – 2002.

Data obtained from Nf.birds Newsgroup 2003). Sightings listed as more than one bird, but not specifying the total number, were counted as 2 birds. Some individuals observed more than once.

Red Crossbills have been seen in a Red Pine stand near Gambo in central Newfoundland. In March 1996, a feeding flock of eight Red Crossbills were observed there during a targeted survey of the stand (Whitaker et al. 1996). No other Red Crossbills were seen from May to October 1996 (Whitaker et al. 1996), but were the most abundant bird species recorded from January to April 1997 (Lewis 1997). A total of 28 Red Crossbills were counted during these surveys (which included a flock of 21 birds during March). Interestingly, Red Crossbills were counted in only one region of the census area, where Red Squirrels were not very common (Lewis 1997).

Field Counts and Typing of Red Crossbills

Estimates of population trends of Red Crossbill percna subspecies based on field counts are complicated by questions surrounding crossbill type identification. Based on records from CBCs, BBSs and other accounts, it is generally not possible to ascertain whether Red Crossbills observed in Newfoundland, or elsewhere, are L. c. percna birds. No quantitative data exist on potential proportions of percna birds found on the mainland, or of mainland subspecies found on Newfoundland. Interestingly, however, of all the Red Crossbill specimens collected on Newfoundland and housed in the Canadian Nature Museum, none have been mainland subspecies (M. Gosselin, pers. comm., 2004).

Parchman and Benkman (2002) contend that “the formerly abundant Newfoundland Crossbill is probably extinct.” This opinion is reflected in previous work by Benkman (1989; 1993c) and Pimm (1990) though no evidence is offered to substantiate this claim. Benkman (1992a) states, “although Red Crossbills are still reported from Newfoundland, most of these are probably other, mainland subspecies of Red Crossbill, which apparently move on and off the island like White-wings.” There is no definitive evidence that indicates that mainland subspecies of Red Crossbill regularly occur on insular Newfoundland (or that they do not).

Some evidence does exist, however, to support the notion that at least some of the Red Crossbills observed recently in Newfoundland are the percna type, and not mainland subspecies birds. A Red Crossbill collected in Terra Nova National Park in 1997 has morphological and plumage characteristics that are similar to those described in the literature for L. c. percna (see Figure 2). Similarly, Red Crossbills that were widely observed around the Memorial University campus in April 2002 exhibited the large bill and dusky plumage characteristic of L. c. percna (see Figure 1). Unfortunately, clear morphological and plumage characteristics of recent Red Crossbills sightings/ collections may not be undisputable evidence of the continued persistence of L. c. percna on insular Newfoundland. Currently, Red Crossbills from the type complexes in North America (Groth 1993b) and Europe (Robb 2000; Summers et al. 2002; Summers and Piertney 2003) are most reliably separated on the basis of call-note typing. There is a dearth of recordings of Red Crossbills in Newfoundland. Groth’s (1993b) call typing was based on only a single recording by Jay Pitocchelli of a Red Crossbill from Newfoundland (Pitocchelli 1981), and current searches have not turned up additional pre-existing recordings, or opportunities to obtain new recordings in the field. When recordings are obtained from mist netted or captured birds in the future, it is important to match calls to bill size (Summers and Piertney 2003).

Conclusions

Red Crossbills in Newfoundland have gone from being a relatively common, breeding, year-round resident during the mid-20th century, to being a species which is rarely encountered today. While the nomadic nature of Red Crossbills makes it difficult to estimate the number of birds in a population, all available evidence indicates that the population size has greatly, and continuously, decreased. It is important to note that the available data set for Red Crossbill abundance (CBCs, BBS, forestry censuses, informal observations, etc.) is concurrently consistent and includes many zero data points. This is not a situation where there is an apparent small population due to a lack of sufficient sampling. 

Estimating the number of percna individuals is extremely difficult. A rough estimate of between 500 – 1500 individuals is suggested, based on field experiences, Christmas Bird Counts, Breeding Bird Surveys and other surveys that suggest that the order of magnitude of the population could be between 100s and low thousands. Much uncertainty is associated with this estimate, for several reasons: 1) observer effort across Newfoundland is relatively sparse and limited; 2) there are no recent observations of breeding so no information on recruitment to the population is available; and 3) the vagility of the birds makes accurate sampling very difficult. Even given the uncertainty surrounding the number of individuals, it is obvious that if the population of percna continues its decline, the taxon will be extinct.