Red Crossbill Research
Conducted by Kenneth Irwin
With the encouragement and generous support of Mad River Biologists, I began an investigation of Red Crossbills (Loxia curvirostra complex). I made most of the recordings included here during the years 2001-2010. There is one exception noted below. I recorded calls from Types 1-5 and Type 10 in Humboldt County, California. Additionally, I recorded Types 2, 5 and 7 near Crater Lake, Oregon in 2005 and one Type 6 crossbill in southeastern Arizona in 2003. I recorded songs from Types 2, 3 and 10 in Humboldt County and from Type 2 near Crater Lake, Oregon.
From 1999 to 2010 more than 11,000 hours were spent observing Red Crossbills. Most of this time was devoted to gathering information on the behavior of Type 10 crossbills in northwestern California with varying periods of time spent with Types 1-5 that were found there as well. Measurements of 89 Type 10 Red Crossbills were made (see Methods in Irwin, K. 2010) and appear below. Details of the behavioral, morphological and ecological differences between Type 10 and other types of crossbills are described in Irwin, K. (2010).
I recorded crossbill vocalizations with a Super unidirectional microphone and a Sony TCM-20DV tape recorder and used Cool Edit 2000 to produce and analyze audio spectrograms from the recordings.
The material herein will be occasionally updated. More songs from Types 2, 3 and 10 will be added in the future.
Click on a folder to expand its contents.Expand All | Contract All
Of the 10 types of crossbills in North America, Type 8 apparently occurred only on Newfoundland; but may now be extinct. Type 9 is restricted to two small mountain ranges in Idaho, and I have no recordings of these two types (for spectrograms of Type 8 see Groth 1993 and for Type 9 see Keenan and Benkman 2008). Type 1 is common in eastern North America with smaller numbers of them in the Pacific Northwest. There also may be Type 1 crossbills in Asia. Although more data is needed, the spectrograms made from 3 recordings on the Borror Laboratory of Bioacoustics website are compelling evidence. On those recordings (made in Sichuan, China) there were 5 or 6 "Red Crossbills". Although one had an unusual high frequency element overlaying the call, which could have been individual improvisation, the spectrograms of the calls from the rest of these crossbills had characteristics that were the same as typical Type 1 flight calls. These 3 recordings were catalogued as Repository ID: 17284; 17287; 17367. In eastern North America Types 1-3 and Type 10 are widespread. Some Type 4 birds have been found in the northern states and recently the first Type 5 crossbills were recorded in the east (Matt young, in review). In western North America Types 2-5 are widespread although Type 3 has a more northern distribution and Type 5 more inland, Type 7 has been found only on the east side of the Cascade Mountains and northern Rocky Mountains, Type 6 in southeastern Arizona (hypothetically in Mexico) and Type 10 occurs near the coast from northern California to the Canadian border (possibly further north).
The following is a description of the occurrence of conifers and Red Crossbills in Humboldt County, northwestern
California 1999-2010. Coast Redwood (Sequoia sempervirens) is the dominant conifer in the fog belt which
extends from the ocean inland for 5-15 kilometers and up to ~ 600 meters in elevation. In this area, Sitka Spruce
occurs in moderately sized tracts. Sadly, the number of Sitka Spruce trees is decreasing. Western Hemlock
(Tsuga heterophylla) is localized with very small patches that are very widely scattered. Shore Pines
(Pinus contorta contorta) are restricted to the area near the coast. Douglas-fir (Pseudotsuga
menziesii) occurs in small numbers at the coast, increases in frequency inland and is the dominant species
from the upper portion of the fog belt eastward. North of California (and Redwoods) Western Hemlock and
Douglas-fir are more numerous at the coast. There are other species of conifers in the fog belt but along with
Redwoods; they do not seem to be utilized by crossbills. At higher elevations inland there are widely scattered,
small to moderate sized outcrops of serpentine. On these there are mixed pine woodlands with Jeffrey
(Pinus jeffreyi), Western White (P. monticola), Sugar (P. lambertiana), here and there
Knobcone (P. attenuata), a few Ponderosa (P. ponderosa) at some sites (much more common eastward)
and moderate numbers of Douglas-fir trees. There are other species as well but they are probably not utilized by
Over most of its range, Sitka Spruce produces cone crops that can vary greatly from year to year (Harris 1990). In contrast, in Humboldt and southern Del Norte counties in northwestern California (from 40.56° to 41.70° N and from 124.08° to 124.35° W) Sitka Spruce produced an abundant cone crop every year from 1998 through 2009. Shore Pines had a moderate crop most years. Although Western Hemlock at the coast had a very poor crop every year (1999-2009), at slightly higher elevations (3-15 kilometers inland) trees had moderate to abundant crops year to year. Douglas-fir trees had poor crops at low elevations and low to moderate numbers of cones above ~ 300 meters.
Benkman (1993) predicted that there might be a form of Red Crossbill with a bill that is shallow in depth and is adapted to Sitka Spruce. Type 10 crossbills were first discovered as they arrived at Patricks Point State Park in November 2001 (Irwin, K. 2010). There were more than 600 Type 10 crossbills that were resident in Sitka Spruce forests 2001-2010. Many flocks of 14-21 Type 10 crossbills each resided in their own relatively small discrete home ranges 2001-2010. The long period of occupancy of sites seemed to be related to the comparatively high level of spruce seeds available year round.
There is growing evidence indicating that the range of Type 10 crossbills extends from northwestern California northward at least as far as British Columbia. North of California a few of them have been found at widely scattered sites in association with Sitka Spruce (Irwin, K. 2010, Thomas P. Hahn, Matt Young personnel communications). In the east, low numbers frequently occur in northern states where they seem to be utilizing White Spruce (P. glauca), Red Spruce (P. rubens), Norway Spruce (P. abies), Red pine (P. resinosa) and Eastern White Pine (P. strobus) (Matt Young, unpublished data).
Although Types 4 and 10 have flight calls that are somewhat similar, there are consistent vocal, morphological, ecological and behavioral differences that are akin to those among other types of crossbills. Type 10 crossbills are much smaller than Type 4 birds. They are intermediate between the smallest, Type 3 and Type 1 and appear to be well adapted for foraging in Sitka Spruce. The medium sized bills of Type 4 birds are better adapted to the stiffer cone scales and larger seeds of Douglas-fir (Benkman 1993), and they seem to prefer that species to Sitka Spruce. The few Type 4 birds recorded in spruce stands remained there only a short time. When both Type 10 and Type 4 were present at a site, they exhibited only minimal social interactions; as has been observed between other types of crossbills.
Benkman (1993, 1999, 2003) and Benkman and Miller (1996) found a positive relationship between the mean depth of the bill of the various call types and the pliability of the scales on the cones of each call type's "key" species of conifer (see figures in Benkman 1993). Crossbills with deep bills are more efficient at opening the cones of conifers with stiff scales than smaller-billed crossbills. Smaller-billed crossbills are more efficient at opening cones with more pliable scales. In the following, the mean bill depths of Types 1, 3 and 4 were determined by Groth (1993a). Type 3 (♂♂ = 8.19 mm; ♀♀ = 8.10 mm) is well adapted for opening the cones of Western Hemlock (Benkman 1993), which has papery scales that are slightly more flexible than those of Sitka Spruce. Type 1 (♂♂ = 8.80 mm; ♀♀ = 8.72 mm) has been hypothesized to be adapted to Red Spruce in eastern North America (Groth 1993a, Parchman et al. 2006). Type 4 (♂♂ = 9.00 mm; ♀♀ = 8.52 mm) is adapted to Douglas-fir (Benkman 1993). Sitka Spruce cone scales are much more flexible than those of Douglas-fir. By comparison, Type 10 crossbills measured in this study (Appendix. Table 1) are smaller than Type 4, and their mean bill depth (♂♂ = 8.55 mm; ♀♀ = 8.37 mm) is intermediate between that of Types 1 and 3. The mean values of bill length, tarsal length and mass for Type 10 (Table 1) are also intermediate between those of Types 1 and 3 (Table 2 and Table 3). The mean wing chord of males and females of Type 10 is slightly shorter than that reported for Type 3 (Table 3).
All Red Crossbills have a groove on the sides of the upper mandible's horny palate and birds hold seeds in this groove while husking them (Newton 1972, Benkman 1993). The width of this husking groove is adapted to the size of the seeds of each call type's "key species" (Benkman 1993). As with some other morphological characteristics, there is overlap in the widths of the husking grooves of Types 1, 3, 4 and 10. Benkman (1993) predicted that a crossbill adapted to Sitka Spruce should have a bill that was shallow in depth and a husking groove ~1.55 mm wide. The widths of the husking grooves of the two Type 10 birds for which I made palate molds (see Benkman 1993) were 1.50 and 1.55 mm, narrower than the mean value of 1.75 mm for Type 4 birds and conforming to Benkman’s (1993) prediction.
From 1999 to 2010 there were 4 major crossbill invasions (1999, 2004, 2006, 2008). Beginning in early October during irruption years, many small sites in Douglas-fir on ridges inland had 100+ Type 2 and 100+ Type 4 and at some sites there were a few Type 3 and Type 5 crossbills as well. It seemed that these birds were moving from east to west. Varying between invasions, small to large numbers of birds of each of these types of crossbills arrived in coastal stands of Sitka Spruce from late October through November. In contrast, during these invasions many groups of 20-100+ Type 3 crossbills (and rarely 1-3 Type 1) arrived in this spruce in October and they seemed to be moving south near the coast. With the exception of Types 2 and 5 at the coast after the 1999 irruption (discussed below), the following pattern was repeatedly observed. Birds remained at all of these sites until the following spring, then almost all of these crossbills left these areas from late May through early June. The exception to this pattern was Type 4 crossbills. They remained at sites in Douglas-fir (inland) but did not stay at sites in Sitka Spruce longer than two weeks. During invasions some crossbills foraged in Shore Pines in tracts on coastal dunes west of Humboldt Bay. Outside of irruption years these pines were seldom visited. At other sites at all times crossbills seldom foraged in Shore Pines that were near Sitka Spruce.
Many Type 10 crossbills nested in Sitka Spruce every year after their arrival (2002-2009). Types 3 and 10 have small bills, which should make them better adapted for foraging in Sitka Spruce than large-billed forms such as Types 2 and 5, which are adapted (respectively) to the stiffer cone scales and larger seeds of Ponderosa Pine and two subspecies of Lodgepole Pine (Pinus contorta latifolia; P. c. murrayana) (Benkman 1993). In May 2000 there were numerous flocks of Types 2 and 3, and two flocks of Type 5 in Sitka Spruce. Types 3 and 5 left late in that month but apparently all the Type 2 flocks stayed and nested in 2000 and 2001. Then they all left in October 2001 as a very good cone crop finished ripening, i.e. at a time when food was at its highest level of abundance. I strongly suspect that the two flocks of Type 5 crossbills that left in May returned in November 2000. Type 5 birds are rare here. The two sites were 2 kilometers apart and the number of birds at each site was roughly the same in November as had been there in May. There was a very small and inconspicuous depression at each site that held water, from which crossbills drank in May, but these were dry in November. Within 1-2 days following their arrival, birds landed next to these depressions and peered intently into them as if they expected to find something, giving the impression that they were familiar with the site from previous experience. Birds in these two flocks nested that winter and these were the only definite cases of winter nesting detected during this study. There also was a flock of Type 2 crossbills found in July 2008. The birds in this flock were the only Type 2 crossbills found near the coast that summer and this was three months prior to the 2008 invasion. There were begging fledglings with these birds, and they foraged in Sitka Spruce, although they may have used another food source as well. This flock remained in a residential area throughout July. The only nesting by Type 3 was by birds in a flock that were frequently eating at a bird feeder (but also spent time in Sitka Spruce).
Because Type 3 have bills that are slightly smaller than Type 10, and hundreds have appeared to do well for many months in Sitka Spruce, I expected to find at least a few that nested using Sitka Spruce as their primary food source. Almost all of these birds left the area late in the spring that followed their invasion and, excluding the aforementioned flock using a feeder, there was no evidence of nesting by Type 3 crossbills. Type 3 appears to be adapted to small-coned hemlocks (Benkman 1993). No crossbill was ever seen in Western Hemlock (although Type 3 does use it further north). Type 4 has been linked with Douglas-fir (Benkman 1993). In this area almost no crossbills were found in Douglas-fir at low elevation and, disregarding observations associated with irruptions, a very few Type 4 individuals were the only crossbills found in forests dominated by Douglas-fir and these were at high elevations. It appeared that crossbills did not nest in Douglas-fir forests. I found two sites in pine woodland (one was near Horse Mountain, west of Willow Creek: 40.8743 degrees N and 123.732832 degrees W; the other was in the vicinity of Onion Mountain, north of Orleans: 41.372901 degrees N and 123.749787 degrees W) where behavior or the presence of fledglings suggested that very small numbers (2 or 3 pairs) of Type 4 birds might have nested. Although there was no direct evidence, I think that in these cases Western White Pine might have been an important food source. I also suspect that a few Type 2 birds might be found in pine woodlands on rare occasions. Additionally, from 1971-2010 I and other observers regularly encountered Red Crossbills in Sitka Spruce but excluding invasions, rarely found them (in small numbers) inland. The actual number of crossbills may have been much greater than these observations indicated. Crossbills can be very easily overlooked. They may remain silent for as much as 2-3 weeks. I have spent days with flocks that I had been studying, when the only indication of their presence was the ‘rain’ of seeds falling that had been dislodged by their foraging activity. Another indication was cones falling. They occasionally pulled a cone off of a tree (which required considerable force), foraged in it and then discarded the cone.
Female Type 10 Red Crossbill.
The recordings here are various calls used by several Types of Red Crossbills (Types 1-7 and Type 10) recorded by Kenneth Irwin. Assortatively, mating groups of crossbills are differentiated primarily by the type of flight call that they use.
Figure 1. Representative sample of spectrograms of the flight calls of North American Red Crossbills. This figure illustrates the modulations of amplitude (thickness of the line) and frequency over time. The Type 5 call is an example of the 2 "voices" giving differentially modulated signals simultaneously. Types 1-7 and 10 are represented (T1-T7 and T10).
There were 7 classes of calls recorded and tracks are labeled using the following system. Two or more tracks were
made for the flight calls of some types of crossbills because including all of the recordings on one track produced a
very large file. The titles of these tracks begin with the type of crossbill recorded, abbreviated as: T 1, etc. Then
the term: flight calls is abbreviated as: F C. The final number differentiates the tracks that have the flight calls
used by that type of crossbill. For instance, T 2 F C 3 is a recording of a Type 2 crossbill giving a series of
unusual flight calls that were discussed in the text and differed from the Type 2 calls on the other two tracks. The
track designated as: T 2, T 4, T 10 Alarm Calls has recordings from three types of crossbills. In the section with
toop calls (or "excitement calls") the title is shortened to Toops. The term Chitters
is likewise used for chitter calls.
Among all of the birds of each type of Red Crossbill the calls differ between individuals and this probably serves to identify each member of a group. The sounds of these different forms of a single type of flight call can differ greatly between the individuals of each type of crossbill. Therefore the best clue for identifying Red Crossbills in the field is to determine the frequency modulation pattern of the flight call.
For more information and recordings see Matt Young's (2008) contribution: Introduction to Differences in Crossbill Vocalizations. This can be found at: http://ebird.org/content/ebird/news/introduction-to%20crossbill-vocalizations and should be updated by the end of April 2010. Also see the Red Crossbill section by Groth on the website: www.amnh.org and see Groth 1993a. For recordings of Type 9 (as well as some other types) refer to: http://www.xeno-canto.org/.
• Frequency Modulation Characteristics
• Typical Flight Calls
• Type 1 Flight Calls
• Type 2 Flight Calls
• Type 3 Flight Calls
• Type 5 Flight Calls
• Comparison of Types 2, 3 and 5
• Type 6 Flight Calls
• Type 7 Flight Calls
• Type 4 Flight Calls
• Comparison of the Flight Calls of Type 4 to Type 10
• Type 10 Flight Calls
• Alternate Type 10 Flight Calls
• Various Alarm and Toop Calls
• Chitter Calls
• Other Classes of Calls
Although it may seem difficult at first, determining whether the frequency rises or falls in at least some portion of a call can be an invaluable aid in identifying Red Crossbills. This first recording is a short duration, rapidly rising call and it is followed by one that has a falling inflection. Then there is a series of each of these, presented in the same order.
Track Title: Comparison of Rising to Falling Inflection
This track has recordings of 8 types of flight calls. The types of calls are arranged in numerical order from Type 1 to Type 7 and then Type 10, as in Figure 1 (no recordings of Types 8 or 9 were obtained). Each recording is a series of flight calls that represent the most typical form of that type of flight call. Among the individuals of each type of crossbill there are many birds that use calls that deviate from the typical form to varying degrees.
Track Title: Types of Flight Calls T 1 – T 7 then T 10
Track T 1 F C 1 is recordings of Type 1 flight calls from some of the few birds found in northwestern California. The staccato sounding calls of Type 1 tend to be slightly shorter in duration than most other flight calls and the frequency falls rapidly. Many are relatively clear sounding.
Track Title: T 1 F C 1
Track T 2 F C 1 is representative of the forms of Type 2 flight calls used by many Type 2 birds. The frequency of Type 2 calls falls slowly. They generally sound deep and full. Some calls have an almost hollow quality. Calls wherein the frequency falls throughout the duration of the call are clear sounding.
Track Title: T 2 F C 1
Track T 2 F C 2 has other forms or variants of Type 2 calls. In these Type 2 flight call variants the frequency first briefly falls then briefly rises before falling slowly and these calls are slightly rough sounding. In this characteristic they are somewhat reminiscent of Type 3 calls, although Type 3 calls fall more rapidly.
Track Title: T 2 F C 2
The last track, T 2 F C 3 is a series given by a Type 2 bird. Spectrograms of these calls look similar to the low frequency element in 27 of the 36 published sonograms of Type 9 flight calls (Smith and Benkman 2007, Keenan and Benkman 2008, Benkman et al. 2009). It appears that the flight calls of Type 9 crossbills tend to be given in a lower frequency band than most other flight calls. The calls on T 2 F C 3 are lower pitched than other forms of Type 2 calls and are reminiscent of one or two recordings of Type 9 calls.
Track Title: T 2 F C 3
Tracks T 3 F C 1 and T 3 F C 2 are recordings of Type 3 flight calls. These calls are characteristically rougher than most of the forms of all other types of flight calls. They are somewhat "scratchy" sounding. The frequency falls rapidly, rises briefly then falls rapidly again. This pattern seems to account for a ch sound. Whereas most North American Red Crossbill flight calls sound like jip, calls given by Type 3 sound a little more like chip.
Track Title: T 3 F C 1
Track Title: T 3 F C 2
Track T 5 F C 1 includes recordings of Type 5 flight calls. These calls are moderately rough sounding and seem to fall slowly. In the Type 5 flight call pattern there are two elements. One is higher pitched than the other and it is delivered simultaneously with the other element. These two elements are differentially modulated and this probably accounts for a resonant quality that many of these calls have.
Track Title: T 5 F C 1
In general, Type 5 flight calls sound slightly rougher than the rough sounding forms of Type 2 calls (track: T 2 F C 2) but not as rough as Type 3 flight calls; and the frequency of the Type 3 call falls more rapidly than Types 2 and 5. The track: Comparison of T 2, T 3, T 5 F C is first a series of slightly rough sounding Type 2 flight calls, then rougher sounding Type 5 calls and then ends with the harsher sounding Type 3 flight call.
Track Title: Comparison of T 2, T 3, T 5 F C
Track T 6 F C 1 is a recording of the only Type 6 crossbill heard. These are reminiscent of some Type 2 calls. The briefly rising portion at the end of the Type 6 call is a distinctive characteristic but may be difficult to hear.
Track Title: T 6 F C 1
Some forms of Type 10 Flight Calls have modulation patterns that are somewhat similar to Type 7 (e.g. Figure 3: #5-#8, #11 and #19) and these variants also have a twang. Type 10 flight calls are characterized by their prominent rising inflection, as are Type 7 calls but Type 10 calls rarely have a final falling inflection and are usually given in a higher frequency band than other flight calls.
Track Title: T 7 F C 1
Some forms of Type 10 Flight Calls have modulation patterns that are somewhat similar to Type 7 (e.g. Fig. 3: #5-#8, #11 and #19 in the file: Type 10 Flight Calls) and these variants also have a twang. Type 10 flight calls are characterized by their prominent rising inflection, as are Type 7 calls but Type 10 calls rarely have a final falling inflection and are usually given in a higher frequency band than other flight calls.
Track T 7 F C 2 is a recording of a series of unusual calls from a Type 7 crossbill that were given as a squirrel ran by. They appear to be flight calls and this may be an example of flight calls being modified and used like chitter calls (see Groth 1993 and Robb 2000).
Track Title: T 7 F C 2
Track: Track: T 4 F C 1 is recordings of many Type 4 flight calls and track: T 4 F C 2 has many calls that are less typical than those on the first track. The frequency falls rapidly then quickly changes and rises rapidly. This modulation pattern apparently imparts a rough quality to most Type 4 calls and some are slightly burry as well. A few calls have a structure similar to that of Figure 2: #3 (in the following section) and these have a twang. In these calls, the twang is nasal whereas in the calls of other types of crossbills that have a twang (Type 7 and some Type 10) it was not nasal. The first recordings on T 4 F C 2 have a nasal twang.
Track Title: T 4 F C 1
Track Title: T 4 F C 2
Audio spectrograms illustrate the similarity in and differences between the flight calls of Type 4 and Type 10.
Figure 2. Comparison of the flight calls of Type 4 with Type 10 Red Crossbills (Irwin, K. 2010). The sample of calls from 5 Type 4 crossbills (#1-#5) represents the range of the variation in call structure between Type 4 individuals recorded in western North America. That is, all of the Type 4 calls began with a falling frequency modulation like that seen in #1-#5 and each had an overall pattern that was either similar to 1 of these 5 spectrograms, or was intermediate in form between typical, like #5 and the most extreme deviation from typical, as #3. The most typical Type 10 flight call pattern (#6-#8) had a frequency modulation that rose throughout at a rate slightly less than the rate (steeper slope) of the rising portion of western Type 4 calls (e.g. #5). They were shorter in duration and were usually given in a slightly higher frequency band.
Figure 2 shows the range of individual variation in the flight calls of Type 4 and provides a comparison of the flight calls of Type 4 with a representative sample of Type 10 flight calls. The most typical flight calls of Type 10 are simple in structure (Figure 2: #6-#8) and resemble the last half or the rising portion of Type 4 flight calls (Figure 2: #1-#5). Type 4 calls recorded in western North America rose more rapidly than Type 10 but in the flight calls of at least some Type 4 crossbills recorded in eastern North America the frequency rose more slowly than type 10 (Irwin, K. 2010). In some Type 4 calls the falling portion of the call is given at low amplitude. In some other variants the part of the call that falls is shorter in duration than typical forms. Distinguishing the falling portion of these forms of the Type 4 call can be difficult. Flight calls given by some Type 4 birds sounded very similar to those given by some Type 10 crossbills.
The track entitled: Compare T 4 to T 10 F C 1 is recordings of Type 4 and Type 10 crossbills whose flight calls sounded similar. Each recording is a single flight call. There are calls from 6 Type 4 crossbills and each of these is paired with a call from one of the 6 Type 10 birds represented. In each pairing the first call is Type 4, then Type 10. The calls deviated slightly from the typical form. The Type 4 calls were higher pitched than normal and the Type 10 calls were lower pitched than normal. Although these two types of crossbills usually sound at least somewhat different, these flight calls were selected in order to demonstrate how similar some calls can sound.
Track Title: Compare T 4 to T 10 F C 1
The track entitled: Compare T 4 to T 10 F C 2 is recordings of more typical calls. In each pairing there is first a series of Type 4 flight calls then a series from a Type 10 crossbill. The Type 4 calls sound rough. In contrast, the most typical form of the Type 10 call is clear sounding, slightly shorter in duration and this as well as other forms of Type 10 flight calls are usually higher pitched than most of the calls of all other North American Red Crossbills.
Track Title: Compare T 4 to T 10 F C 2
Type 10 flight calls have a strong, rising inflection. One of the best clues for distinguishing Type 10 is that most of their flight calls are higher pitched than most of the flight calls of all other North American Red Crossbills.
Figure 3 (3a, 3b, 3c). Flight calls of 35 Type 10 Red Crossbills. Figure 3 represents the range of variation between individuals in the Type 10 population. Each one of the first 3 sonograms (#1-#3) is representative of many individuals that had a call similar to it. #4, #5 and #6 each represent the flight call of a more moderate number of individuals. Each one of the other modulation patterns (#7-#35) is representative of fewer individuals whose call was similar to it. The diversity of forms indicates that the range of individual variation among the Type 10 population is greater than any other type of Red Crossbill.
The structure (and sound) of the Type 10 flight call can vary greatly between individuals (Figure 3). The most typical forms of the Type 10 flight call are simple in structure and the frequency rises throughout the call (Figure 3: #1-#4). These forms are clear sounding. Other forms can sound rough. Some forms of Type 10 flight calls have a modulation pattern wherein the frequency falls, then rises after a long transitional period (e.g. Figure 3: #5-#8, #11 and #19). This is reminiscent of the Type 7 pattern in Figure 1 (below). Calls with forms similar to this have a quality that can be likened to a slight twang that is not nasal. This can sound vaguely reminiscent of the twang in Type 7 calls but Type 10 calls are higher pitched. In some other forms of Type 10 flight calls the change from falling to rising (and rising to falling) frequency is more rapid (e.g. Figure 3: #12, #13, #24 and #35) and these kinds of variants are burry sounding.
Track T 10 F C 1 is recordings of birds using Type 10 flight call variants that are similar to the most typical form of the Type 10 call and each is representative of many individuals that used a similar call (Irwin, K. 2010).
Track Title: T 10 F C 1
Track T 10 F C 2 is recordings of two forms of Type 10 calls and each is representative of many birds that had a similar call.
Track Title: T 10 F C 2
Track T 10 F C 3 is recordings of other forms of Type 10 calls and each is representative of a more moderate number of birds. The track begins with two recordings of calls that have a twang.
Track Title: T 10 F C 3
Tracks: T 10 F C 4 and T 10 F C 5 are recordings of other variants that are rather unique and each one is representative of only a few Type 10 individuals that used a flight call that was similar to it.
Track Title: T 10 F C 4
Track Title: T 10 F C 5
Many Type 10 crossbills used two different forms or variants of Type 10 flight calls.
Figure 4. Individual Type 10 Red Crossbills that used two different forms of Type 10 flight calls. There are 6 birds and the 2 calls that each one used. Each individual used both a call with a simple modulation pattern (e.g. #1) and a complexly modulated call (e.g. #2).
In the recordings the differences may not be aurally distinguishable but the disparities are evident in spectrograms. Most of the birds that had two variants used one form that was simple in structure (e.g. Fig.3: #1-#4) and a more complexly modulated form similar to one or another of the forms represented by Fig.3: #5-#35. It seemed that individuals used the two forms differentially (discussed in Irwin, K. 2010).
Track: T 10 Alternate F C 1 is recordings of two calls from each of 6 Type 10 crossbills. There is first a series of one call then a series of another Type 10 flight call variant that the individual used at a different time. In each case the first call is simple in structure and the second call in the pairing is more complexly modulated. The spectrograms in Figure 4 were made from the calls on this track and are depicted sequentially in the same order as the calls presented on this track.
Track Title: T 10 Alternate F C 1
Tracks T 10 Alternate F C 2 and T 10 Alternate F C 3 are recordings of 28 Type 10 crossbills, each voicing a series of flight calls. In each case the bird changed from using one Type 10 flight call variant to its alternate Type 10 flight call variant during the short series of calls.
Track Title: T 10 Alternate F C 2
Track Title: T 10 Alternate F C 3
Alarm calls were used briefly and were seldom heard. They were low frequency calls and were given at low amplitude. Alarm calls were often given in a halting delivery rate.
On the track: T 2, T 4, T 10 Alarm Calls, the first recording is a series of three calls given by a Type 2 crossbill, then three by Type 4, and the last recording is two alarm calls by Type 10.
Track Title: T 2, T 4, T 10 Alarm Calls
Although Nethersole-Thompson (1975) described a "toop call" given by Scottish Crossbills (L. scotica) associated with anger, alarm or excitement, Growth (1993) called this class of calls "excitement calls". For each type of crossbill, it appears that the alarm call and the toop call are related because toops can be used for alarm and the two calls are similar in structure (see Growth (1993). Toops are given in a frequency band that is slightly higher than alarm calls and are usually given at the same rate as flight calls. Toops are used in association with alarm, excitement, aggression or reproductive behavior.
Type 1 toops are slightly modulated. The frequency rises moderately then falls briefly. They can sound like some toops given by Types 2, 4 and 10.
Track Title: T 1 toops
Type 5 toops are short in duration and tonal.
Track Title: T 5 toops
Type 3 toops have two prominent components. One "voice" gives a low frequency element that is similar to a Type 4 toop (e.g. the low frequency element in Fig. 5: #2, below) while the other "voice" simultaneously delivers a high frequency element that rises then slowly falls and these two modulation patterns together produce a dissonance. The nasal sounding toops of Type 3 crossbills are distinctive.
Track Title: T 3 toops
Figure 5. The toop call (or "excitement call") of Types 2, 4 and 10 Red Crossbills. Type 2 is represented by #1, Type 4 by #2-#4 and Type 10 by #5-#8. Some toops were so similar that it was difficult to differentiate between these three types in a few cases. Many toops from Types 2 and 4 had harmonics. Type 10 crossbills had a strong tendency to include a signal produced by the high voice that overlied the low voice and continued long after the low voice terminated (#5-#7). Toops that were temporally prolonged (#4) were occasionally given by Type 4 crossbills. Type 10 birds also used an alternate form of toop that was prolonged (#8) but this form was unlike Type 4 toops.
The toop calls of Types 2, 4 and 10 were similar. The toops of Types 2 and 4 often had harmonics. Harmonics were seldom recorded with Type 10 toops.
Type 2 toops have a rising inflection.
Track Title: T 2 toops
Type 4 toops also have a rising inflection. It appears that Type 4 crossbills may have two different forms of toop calls. The short form (Figure 5: #2) can sound the same as Type 2 toops and some Type 10 toops. Some Type 4 individuals used toops that were unique in structure. They were longer in duration; more modulated, and had a greater frequency bandwidth than others (Figure 5: #4). The sound of this call was distinctive and unlike any other toop. The first 6 recordings on the track with Type 4 toops (0-11.5 seconds) are birds giving the short form of the call and the next 4 recordings are birds using the sweeter sounding long form. The final recording is a long series wherein the calls grade from the short form to the long form.
Track Title: T 4 toops
Type 10 crossbills used toops that consisted of a low frequency element given simultaneously with a high frequency signal that overlapped the last portion of the low frequency element. The structure of the high frequency element was usually reminiscent of a first harmonic. Type 10 crossbills tended to continue this high frequency element for a long period of time after the low frequency element had terminated. This produced a hint of a falling inflection at the end of the initial rising inflection and this was detectable in many Type 10 toops. The inclusion of this component made many Type 10 toops slightly longer in duration than the toops of other types. Type 10 crossbills sometimes used a prolonged toop that was different than the toop of any other type of crossbill (Figure 5: #8). It was given occasionally by one or two birds that were in groups that were tooping at Pygmy Owls (Glaucidium gnoma) and was also used in the context of reproductive behavior. On the track with Type 10 toops the first 7 recordings are birds giving the short form of the call. Following those there is a recording of a prolonged form with a strong, falling inflection that was given in response to a Pygmy Owl. Then the track ends with a different, more complexly modulated version of the prolonged form that was given by a female prior to copulation.
Track Title: T 10 toops
Chitter calls (see Groth 1993, Robb 2000) are weak and were infrequently heard. They seemed to be associated with a moderate level of fear. They were most often heard as birds flew down from the tops of trees to bushes next to ponds where they drank. In this situation birds were potentially more vulnerable to predation and they were always skittish or nervous (Irwin, K. 2010).
Figure 6. Chitter calls of Red Crossbills. Type 2 (#1 and #2), Type 3 (#3-#5), Type 4 (#6-#9), Type 10 (#10-#14). Note that the initial long falling portion of Type 4 calls appear concave whereas in Type 10 calls the initial falling portion appears convex. Then Type 10 calls rise more rapidly than that portion of the Type 4 calls.
Chitter calls were recorded from Type 2 (track: T 2 chitter), Type 3 (track: T 3 chitter), Type 4 (track: T 4 chitter), and Type 10 (track: T 10 chitter). Figure 6 illustrates the differences in the chitter calls of these 4 types. The chitter calls of other types of North American crossbills have not yet been reported. The chitter calls of each of these 4 types of crossbills differs from the chitters of the other 3 types. Although the chitter calls of Types 4 and 10 appear similar there are consistent differences. In the initial long falling portion of Type 4 calls the rate of frequency change slowly increases whereas in Type 10 calls the rate decreases. Then the Type 4 call rises at a constant rate that is lower than the rate in the rising portion of the Type 10 chitter call. The differences among these four types indicates that each type of crossbill may have a chitter call that is unique. Robb (2000) also suggested that the chitter calls of types of crossbills in Europe were likewise differentiated. Flight calls and possibly toops that have been modified to varying degrees can be given like chitters (noted also by Groth 1993, Robb 2000).
Track Title: T 2 chitter
Track Title: T 3 chitter
Track Title: T 4 chitter
Track Title: T 10 chitter
Whereas the flight, alarm, toop and chitter calls differ between types of crossbills, each of these three kinds of calls is undifferentiated among the many types of Red Crossbills that share them. The structure of the Begging Calls apparently does not differ between any of the types in North America and Europe including Scottish Crossbills (L. scotica) and Parrot Crossbills (L. pytyopsittacus) (see Nethersole-Thompson 1975, Groth 1993a, Cramp and Perrins 1994, Robb 2000). The Aggression Calls were used during fights. Type 2 and Type 10 used the same call. In Europe Robb (2000) incidentally recorded it used by Type A and Parrot Crossbills. When birds were startled they often used the 'fright' call. Types 2, 3, 4 and 10 all used Fright Calls that had the same structure. All of the calls on all of the tracks in this section were given by Type 10 crossbills.
The begging call has been referred to as the "chit-too call". Some may sound more like: chee-choo. There is a short pause between the two components. It seems that two to three weeks after fledging the begging calls change somewhat and the second component is repeated 1-3 times. The third (and last) recording on the track: T 10 Begging Calls is an example of this.
Track Title: T 10 Begging Calls
Birds that were fighting gave the aggression calls that were recorded. In this call the frequency falls rapidly then quickly changes and rises rapidly then falls again and this pattern is repeated continuously without a break producing a moderately long buzz. In the middle of the first fight on the track: T 10 Aggression Calls one bird stops calling and the other starts.
Track Title: T 10 Aggression Calls
Birds that were foraging on a cone or were perched and then were startled as they were supplanted by another crossbill frequently gave the 'fright' call. This call is a rapidly falling element that is repeated 2-5 times in quick succession, producing a short, high frequency buzz. The track: T 10 Fright Calls is Type 10 birds that were recorded as they where being displaced.
Track Title: T 10 Fright Calls
Male Type 10 Red Crossbill.
In 1999 I set out to learn to identify Red Crossbills (Loxia curvirostra complex) in the field. This took longer to accomplish than I initially anticipated. During this time I noticed what I thought was unusual vocal behavior that raised some intriguing questions. While attempting to resolve these questions I discovered a new type of crossbill (Type 10) in Humboldt County, northwestern California that apparently are well adapted to the cone and seed characteristics of Sitka Spruce (Irwin, K. 2010).
Groth (1993a), Benkman (1993), Smith and Benkman (2007) and Snowberg and Benkman (2007) have suggested that at least some types of Red Crossbills have evolved through a process of divergence with gene flow or sympatric speciation. Smith and Benkman (2007) and Snowberg and Benkman (2007) suggest that for some types of crossbills a divergence in morphological adaptations to optimize feeding efficiency in a new ecological system could lead to habitat divergence and that this differential habitat use can lead to assortative mating through a process of ecological speciation (Orr and Smith 1998, Rundle and Nosil 2005).
There is growing evidence that at least some types of crossbills represent biological species (Groth 1993a,b, Smith and Benkman 2007, Benkman et al. 2009). Benkman et al. (2009) have argued that Type 9 represents a biological species on the basis of morphological and genetic divergence from other sympatric types of crossbills and high levels of reproductive isolation.
Two or more types of Red Crossbills often nest in the same area and mate assortatively by flight call type (Groth 1993b, Edelaar et al. 2004, Smith and Benkman 2007, Summers et al. 2007). In one investigation < 1% of the pairs identified were composed of two types (Smith and Benkman 2007). Flight calls play a role in assortative mating. Snowberg and Benkman (2007) conducted mate choice trials and found that ♀♀ preferred ♂♂ that had the same type of flight call as their own. This suggests that sexual imprinting on this marker trait may be involved in mate choice. Sexual imprinting could play an important role in the evolution of assortative mating (Irwin, D. E. and Price 1999) in diverging groups udergoing ecological speciation. In the field different types also flock assortatively (Groth 1990, Smith et al. 1999, Thomas P. Hahn, personal communication). Adkisson (1981) found that Pine Grosbeak individuals preferred birds that had the same type of location call as their own and these populations flocked assortatively. During this study many observations were made at sites where from 2 to 5 types of crossbills were present. These birds sorted into groups by flight call type and the most obvious cue (often the only apparent cue) was the flight call. Red Crossbills obtain mates from within their own flock (Newton 1972, Smith et al. 1999) as do some other cardueline species (Newton 1972). In this investigation, excluding the movements of juveniles, intensive monitoring of Type 10 flocks indicated that these individuals had a high degree of fidelity to their flock (Irwin, K. 2010) and presumably most matings were between flock members. Although there often were many Type 3 birds nearby during the formation of pair bonds, no pairing by two types was detected in this study.
Little is known regarding how calls changed as groups diverged, leading to new types of flight calls. Further work on vocal behavior should advance our understanding of the evolution of these fascinating birds.
Under normal conditions only a small portion of a group of crossbills vocalizes at any given time. For instance, flocks are usually composed of ~18 birds and when these fly there normally are only 2-5 birds in the flock that call. During irruptions when flocks first arrive at a site it seems that almost all of the birds call frequently. This gives the impression that they are excited. After they have been at a site for three weeks the rate of calling diminishes. At times crossbills can remain silent for very long periods. I have observed flocks for 6-7 hours every day and many times did not hear anything for as much as 2-3 weeks. A lot of these observations were made at sites where I could have heard a second large flock nearby. Lengthy periods of silence were more common in the winter than the summer. Red Crossbills also sang sporadically and often did not sing for extended periods of time (discussed below).
Assortatively mating groups of crossbills are differentiated primarily by the type of flight call they use.
Figure 7. Representative sample of spectrograms of the flight calls of North American Red Crossbills. This figure illustrates the modulations of amplitude (thickness of the line) and frequency over time. The Type 5 call is an example of the 2 "voices" giving differentially modulated signals simultaneously. Types 1-7 and 10 are represented (T1-T7 and T10).
Although the various types of flight calls are readily distinguishable by examining audio spectrograms,
determining whether a vocalization is a flight call or a song component can be problematic because some song elements
are similar to flight calls (discussed below). It may be necessary to observe the bird for an extended period of time
to determine whether it is singing or giving flight calls.
Certain song components are so distinctive that they may be indicative of the type of crossbill that is singing. Other song elements can be misleading. At least some types of crossbills have elements that are similar to types of flight calls that are different than their own flight call type (see: Stereotyped Songs that Seem Like Mimicry).
Recordings of songs are labeled in a way that differs from the system used for calls. For instance, in the section: Distinctive Songs, there is a Type 2 song used during the song bout on track T 2 Song A 2. An excerpt of this was used to create track: Distinctive T 2 Song A - 1, which is used to exemplify the song discussed in the text. The letter A denotes an individual, the dash preceding the final number indicates that the track is a segment of a longer recording and the number specifies the track. Conversion to mp3 format to facilitate accessibility has degraded some portions of these recordings to varying degrees.
Some characteristics of Type 10 vocal behavior differ greatly from the behavior of other types of Red Crossbills. For an analysis of Type 10 calls see Irwin, K. (2010). The most typical flight calls of Type 10 have frequency and amplitude modulation patterns that are very similar to the second half of Type 4 flight calls but the rates of frequency change differ consistently and they are given in a higher frequency band than Type 4. A very large proportion of Type 10 crossbills use less typical forms or variants of the flight call with structures that differ greatly from all other known North American crossbills and there is a very broad range of variation among Type 10 individuals. Many Type 10 crossbills used two different forms or variants of Type 10 flight calls (see: Alternate Type 10 Flight Calls). Of these, most used a simply structured call and an alternate call that was more complexly modulated. It seemed that birds that used two variants tended to use a simple call for communicating with other members of their group and a complex call for communicating with their mate or in the context of reproductive behavior (Irwin, K. 2010). The use of a complex call by Type 10 crossbills probably enhances recognition of individuals and mates because they differ between individuals to a greater degree than simple calls do.
There are potential pitfalls in identifying Red Crossbill flight calls. There also may be problems in identifying a crossbill that is singing if that type’s song repertoire is not known. Birds will often fly to a perch, sing occasionally (for up to an hour) and then fly off without giving a single diagnostic call (flight call, toop, etc.). Although Red Crossbills may occasionally use their flight calls in their songs, more often they only give flight calls during breaks between songs. These pauses can last from ~700 milliseconds to 10 minutes or more. If the intermission is short and the bird voices a series of its flight call during the break it can seem as if the flight calls are an integral part of the song. Another factor is that some types of crossbills have stereotyped song components that are similar to their own type of flight call. Type 10 birds use elements that are very similar to several Type 10 flight call variants. This appears to be the case with Type 2 as well. Moreover, at least three types of crossbills each have two or more stereotyped song elements that are similar to types of flight calls that are different than their own flight call type.
In the discussion below, the term "element" refers to a single, unbroken vocalization. The term "unit" is used for a stereotyped group of elements and in most of these different units the elements in each one were delivered so rapidly in succession that each unit usually sounded like a single note or syllable.
Figure 8a, b. Flight calls and Type 10 song elements that are similar to them. The upper half of the figure is the typical form of the flight calls of Types 1-7 and Type 10 (as in Fig. 7). The lower half is Type 10 song elements that are stereotyped components of their repertoire that are similar to the flight calls of other types of crossbills.
When singing, birds occasionally used vocalizations that were similar to the flight calls of other types of crossbills. The Type 10 stereotyped song elements that are similar to the flight calls of other types of crossbills are illustrated in Figure 8. Most of these patterns have been found in the stereotyped repertoires of Type 2 and Type 3. Both Type 2 and Type 3 have an element that is virtually the same as the Type 10 call (T 10) in Figure 7. Although Type 2 have a Type 5 like element, Type 10 do not and in the moderate sample of Type 3 songs examined there was nothing found that was similar to a Type 5 call.
The possibility that these were cases of mimicry, or birds that had learned the flight calls from other types of crossbills
and were imitating them was considered because it is well known that some cardueline finches imitate the calls of
other species and use them while singing. In the case of many Type 10 birds; it appears that these song components
were learned from adult Type 10 males acting as the models and not individuals of another type. Some of these
elements are similar to the flight calls of types that were not present in the area at the time. In most years, no
individuals of Types 1, 2 or 4 were found in the area that Type 10 resided in; and Types 6, 7 and 9 have never been
recorded in northern California.
Another argument against mimicry is that these song elements are stereotyped and vary little in structure. In contrast, the flight calls of every group of crossbills vary structurally between individuals (see figures in Groth 1993a, Irwin, K. 2010 and Figure 3 in: Type 10 Flight Calls). If these birds were imitating the calls given by birds of another type, then some reflection of this individual variation in flight calls would be expected when comparing the songs from different birds. No evidence of this was found in the spectrograms. Furthermore, crossbills have used these like other stereotyped song constituents to construct novel songs and flight call variants. Type 10 Red Crossbills occasionally improvise new song units by combining two or more song elements. This spontaneous combinatorial improvisation is short lived or ephemeral but provides insight into one process that Red Crossbills use to create new vocalizations. Several integral components of the Type 10 song repertoire were apparently created in this way.
Certain "multitasker" elements have, in various cases, been combined with 1-4 song elements to create 2, 3 or 4 new song elements or units that have become stereotyped. Red Crossbills have apparently used some of the song components that are like flight calls to improvise several elements, units or phrases by combining each one with one or more different song elements and these vocalizations have become stereotyped components of the respective song repertoires.
Type 10 crossbills have a phrase that ends with an element that is similar to a typical Type 2 flight call and use a unit that is composed of an element that is immediately followed by the same Type 2 like pattern. They have what looks like a slightly modified Type 3 flight call followed by a variant of a Type 10 flight call. They also have an element that begins with a Type 4 pattern then ends with a long, rapidly falling component. Type 3 birds use a unit that begins with a Type 2 pattern and have a phrase with the same Type 2 pattern included with different elements. They also use a complexly modulated element that ends with a Type 4 pattern. Type 2 crossbills have an element that is like a slightly modified Type 4 pattern with a Type 1 pattern added to the end.
The song repertoire of Type 10 crossbills in northwestern California 2001-2010 is better known than any other group. They have 6 elements that are similar to 6 types of flight calls (Figure 2). Crossbills will, at times, sing very simple songs comprised of only a single element repeated at the same rate as flight calls. It is therefore possible, on rare occasions, to find a bird giving what seems to be a series of typical flight calls of a different type of crossbill. Early in this investigation I encountered a bird that I thought was a Type 5 crossbill giving flight calls. After 3 minutes the bird flew off using Type 2 flight calls. In this case the Type 2 crossbill was probably singing, simply repeating a single, often used stereotyped element that is similar to a Type 5 flight call, which is an integral component of the Type 2 song repertoire. The track designated as: T 2 Song Like T 5 Flight Calls is a series of Type 5 flight calls then a series of this song element.
Track T 10 Song Like T 2 Flight Call is first a frequently heard, short Type 10 song phrase that ends with an element that is similar to a typical Type 2 flight call. Occasionally birds will omit the first few elements and give a series of only the last element (see: Stereotyped Variations of a Phrase). The second recording on this track is an example of this abridged version of the song and it is followed by a series of Type 2 flight calls.
The structure of the third element in the Type 3 phrase on the track entitled: T 3 Song Like T 2 Flight Call is virtually the same as the last element of the Type 10 phrase on T 10 Song Like T 2 Flight Call, although it is given in a slightly higher frequency band.
Track Title: T 3 Song Like T 2 Flight Call
Track T 2 Song Like T 7 Flight Call is a highly stereotyped element from the song repertoire of Type 2 crossbills then Type 7 flight calls.
Track Title: T 2 Song Like T 7 Flight Call
The track: Variations of a T 10 Song A 1 is recordings of stereotyped components of the Type 10 repertoire that are variations of one motif. The stereotyped variation is in the form of the last element of the phrase. All of these song elements are similar to variants of Type 2 flight calls. There may also be individual variation, although it is usually short lived. For instance, birds may omit some or all of the first parts of the phrase. The first recording is what is thought to be the “classic” rendition of this motif and in this variation the form of the final element is virtually identical to the most typical form of the Type 2 flight call. This variant is rather seldom heard. Birds usually omit the initial high note. The second recording is the rendition most often heard and is the same as on the track entitled: T 10 Song Like T 2 Flight Calls. The third recording is unusual. In this case the bird gave a third variation of this motif and then changed to a stereotyped fourth variant different than others on this track. Birds seldom used two variants in one bout of singing. The last recording is a fifth variant of this phrase.
Track Title: Variations of a T 10 Song A 1
These song elements similar to the flight calls of other types of crossbills may have originated as an imitation of another
type of crossbill's flight call by a bird that continued to utilize it and then was copied and imitated by others in
its group; eventually becoming a stereotyped component of that group's repertoire. During all of the time spent
listening to crossbills and examining recordings nothing was found that suggested that Red Crossbills imitated a
vocalization made by any other species of bird.
If Red Crossbills imitate other types of crossbills they seem to be very selective of what they imitate. Examination of recordings indicated that a very few Type 10 crossbills may have briefly imitated the flight calls of other types of crossbills and no other vocalization. Four individuals that were singing each gave only one single, short series of a vocalization that looked as if it could have been an imitation of a flight call (although some could have been miscopies of stereotyped song elements). Three of these series were similar to Type 4 calls. The fourth bird gave a series of 5 vocalizations. One was the song element similar to a Type 3 flight call, three were obvious miscopies and the fifth might have been an imitation of a Type 3 crossbill's flight call, although it differed from most of those calls.
From 2001 through 2010 most Type 10 crossbills seldom (or never) encountered Type 4 in small numbers and only spent a relatively short period of time near Type 4 birds. In contrast, many large flocks of Type 3 birds spent 7-8 months near many Type 10 flocks on three occasions and were found in smaller numbers for varying periods at other times as well. For these Type 10 crossbills, the flight calls of Type 3 were one of the most frequently heard sounds but apparently almost all Type 10 individuals never imitated Type 3 calls.
Some Type 10 individuals improvised their flight call by adding a song element.
Among the birds of each type of Red Crossbill, the structure of the flight call is relatively labile; varying between individuals. This probably helps them to identify each bird in the group (Mundinger 1970, Keenan and Benkman 2008, Sewall 2008). Among Type 10 crossbills there is a very broad range of individual variation in flight calls (Irwin, K. 2010 and see Figure 3 in: Type 10 Flight Calls). The most typical calls are simple in structure (as is T10 in Figure 1), but many have more complex amplitude and frequency modulations. Sewall (2008) found that when Red Crossbills first develop their flight call, some birds improvise their call by combining portions of flight call variants given by two other members of their family. At least 23 Type 10 individuals developed new flight call variants by combining different song elements with a Type 10 flight call, producing a more complexly modulated compound call. Many of these can be found in Figure 3. Mated pairs usually have very similar flight calls ("call matching"). One bird may imitate its mate's call, or each adjusts its call such that they both converge on a call with a form that is intermediate between the two original variants (Mundinger 1970, 1979, Groth 1993b, Keenan and Benkman 2008, Sewall 2008). At least 6 of the calls that were modified by the addition of song elements were shared by a bird and its mate that were call matching; demonstrating that these were recognized as flight calls (Irwin, K. 2010).
Figure 9 (a and b). Type 10 flight calls altered by the addition of song elements. Figure 9a depicts flight calls that have been combined with song elements and 9b illustrates the song element (elements) within each call after omitting the original flight call structure. #12 is entirely a song element that was used as a flight call (see text).
Eleven of these flight call variants are depicted in Figure 9a. Each spectrogram in Figure 8b illustrates the song elements used in Figure 9a with the original flight call structure omitted. There is one exception. The last flight call; #12 is similar to a song element in its entirety, with only a minor modification of the initial amplitude modulation. Although superficially similar to some complexly modulated Type 10 flight calls, the rates of frequency change or slopes of the first two modulations (first rising then falling) differed from all other Type 10 calls. This call was used by a mated pair and their offspring and was aurally very similar to many Type 10 flight calls. The #1 individual used a complexly modulated flight call given at a relatively lower frequency and simultaneously included a song element that was given in a higher frequency band. In #2, the bird used a song element similar to a Type 1 flight call (see Figure 8) that preceded a typical Type 10 flight call with simple structure. In #5 it appears that an element with a pattern like a Type 2 flight call was added to the end of a greatly simplified Type 10 flight call. In #6 there was a song element that preceded a flight call and together they produced a call reminiscent of a Type 4 flight call with a short silent pause in the middle (there is also a break in some Type 4 flight calls). In #8, the bird gave a typical Type 10 flight call with a song element with a pattern like a Type 1 flight call added to the end, and the bird included an overlying song element at the beginning of the call. The four additions to the typical flight call by the #10 individual made this the most intricately modified flight call found. In #11 the two song elements combined with the complexly modulated call are the most utilitarian "multitaskers" in the Type 10 song repertoire. Either singly, or in the combined form (unit) illustrated here they have been used to create several stereotyped song units. This bird also used an alternate flight call that had a simple structure (Figure 4: #11 and #12) and his mate precisely matched both of these.
Most of the following discussion is based on 10,000 hours of observing the behavior of Type 10 Red Crossbills that
inhabited coastal Sitka Spruce forests in northwestern California, and analysis of the recordings made from 2001 to
2009. Additionally, observations of other types of crossbills were made there and in other areas as well.
During 4 irruptions, numerous crossbills (Types 1-5) invaded these spruce forests in the fall and left the following year; usually late May to early June. During these invasions the number of Type 3 crossbills was much greater than all other types and there were at least a few Type 3 birds that wintered every year. The great majority of the birds in these 5 groups did not sing during their stay. I suspect that after leaving, individuals of each of these types of crossbills waited until they found appropriate conditions for nesting and then began to sing.
In contrast, resident Type 10 crossbills sang more frequently than the other types observed in Sitka Spruce forests. Although rare, songs were heard following the fall molt (October to December). In most years singing behavior increased very slowly from December to April, increased more rapidly through April and May, then diminished through August and September. This tendency to sing more than other types in Sitka Spruce forests may have been due to their being better adapted than other types for foraging in Sitka Spruce, and to their nesting within their year round home range, which provided them with abundant food. They sang more frequently when other types were present. Singing behavior was infectious for all types of crossbills. If three or more crossbills were singing it seemed to motivate others to sing.
The Type 10 stereotyped song repertoire included many different elements. A few songs were used only rarely. Although many males sang frequently in 2002 and 2003, I continued to record components of their repertoire that were new to me until June, 2003. Because no new stereotyped songs have been recorded since then, I believe that the entire repertoire used by Type 10 in northwestern California from 2002 to 2010 was documented. It is not known whether or not there is geographic or intergroup variation (e.g. dialects) in song.
In late June and July of 2006, many songs given by Type 2 crossbills were recorded at scattered sites ~30 kilometers east of Crater Lake in central Oregon. A few brief songs were recorded from one or two Type 4 birds at one of these sites. Through May and early June 2007 many songs from 5-10 Type 3 birds in one flock were recorded in northwestern California; a rare occurrence. There was no singing heard from the many other Type 3 flocks monitored that year. Then, all of the Type 3 crossbills suddenly left. Although a great amount of time and effort was put into trying to record the songs of various types of crossbills, I have what are presumably only moderate portions of the repertoires of Type 2 as well as Type 3 and a small sample of Type 4.
The songs of Types 2, 3, 4 and 10 differ greatly. I suspect that the repertoires of Types 2 and 3 may be as diverse as Type 10; suggesting that all types may have complex repertoires, and that the repertoires may differ greatly among all of the types of crossbills. There was no discernible change in the song elements or the repertoire that the group of Type 10 crossbills in northwestern California used from 2002 to 2010; suggesting that for this group, song evolution was a slow process. I suspect (as does Thomas P. Hahn: personal communication) that this condition may be unusual and that song evolution in other groups or other types may be more rapid. Birds may miscopy song components that they hear and try to imitate. These errors in copying can pass through a group and accumulated copying error plays an important role in the evolution of song. Over the relatively short duration of this study there was no apparent accumulation of copying error. A few residents developed a slightly different variant of one song or another but as yet only one or two birds have imitated these variations. In other species, different song variants (due to differentiation between subpopulations or dialect groups) that are introduced into a population by birds that move into an area can help to drive song evolution. There has been no evidence of Type 10 birds moving into this area since the initial invasion in 2001 (and, as yet, dialects are not known to occur).
The nesting success of Type 10 crossbills was good every year from 2002 through 2009. Although many birds sang at a relatively high amplitude level, and sang frequently in 2002 and 2003; in 2004 almost no songs were heard. In 2005 and 2007 most songs were given at low amplitude ("whisper song"). In 2006, 2008 and 2009 singing was infrequent and many songs were given at low amplitude.
Crossbills tend to flock assortatively by flight call type. Like some other cardueline finches (Newton 1972), it appears that most crossbills obtain their mates from within the flock that they winter with (Newton 1972, Smith et al. 1999). Type 10 individuals apparently had a high level of fidelity to their flock (Irwin, K. 2010) and therefore, during this investigation most of the pairings were presumably by birds that had been familiar with each other for a long period of time. Since males were usually in close proximity to their prospective mates, it may not have been necessary to sing loudly to attract a mate. This may, to some extent; help to explain why Red Crossbills often sang at relatively lower amplitude levels. Additionally, "whisper song" would expend less energy and might draw less attention from predators. Type 10 males did not sing frequently on territories as many other species do. Although the exact circumstances were not clear, a few observations suggested the possibility that some crossbills may have maintained a rather loose pair bond through the fall and winter, which might have diminished the motivation to sing. This may partially account for the great reduction in singing during 2004, as compared to 2002 and 2003.
When a male chased a female he usually sang during the flight. Songs were also given during the flight displays known as "Moth-flight" and "Butterfly-flight" (see Cramp and Perrins 1994). During this investigation Moth-flight was seen occasionally but Butterfly-flight was rare.
The following is based on a great many observations, but the evidence is anecdotal and these interpretations are speculative at best. Although there was much variation by and among Type 10 individuals, there seemed to be some tendencies in singing behavior. Songs heard early in the reproductive period were usually composed of elements and units that were comparatively simple in structure, short in duration and relatively clear sounding. These were the kinds of songs that were used during chases and flight displays and it seemed as if they were primarily used for attracting a mate.
In general, as the reproductive season progressed; birds increasingly used song
elements, units or phrases that were variously buzzy or raspy sounding.
Due to their increased usage later in the season and the ways in which birds tended to use them, it seemed that it
was possible that the use of these kinds of songs increased following the formation of the pair bond. Mated pairs
often moved away from all other crossbills for long periods of time. During these periods, when the male of one of
these pairs heard another crossbill calling from a perch nearby, or flying in its general direction it very often
responded by singing. If another male began to sing, these males usually sang in response. In both of these
situations, buzzy and raspy components were prominent constituents of songs.
If one male responded to the singing of another male; this immediately developed into a bout of matched counter-singing wherein each matched any change made by the other bird. No bird ever approached a male that was giving buzzy or raspy songs. It appeared that for these crossbills, as with some other species, counter-singing was linked with spacing of the singing birds. It seemed that songs with buzzy and/or raspy components could have been used to discourage intrusion or maintain some distance between a pair and other crossbills. Although a few of these were rather short in duration, many were much longer than most other song components (e.g. 300-550 milliseconds vs. ~60). A few of these song components were trills but most were a single, continuous element.
Variously, the fundamental structure (as well as the sound) of many of these was similar to the aggression call, which was undifferentiated among types and used by many (if not all) types of crossbills during fights. The track that is entitled: Aggression Calls and Similar Song Elements has comparisons of aggression calls to buzzy song elements.
The first, third and fifth recordings are aggression calls. The second, fourth and sixth are each a stereotyped song element that is distinctly different from all others. Each song element is presented after a brief pause that follows an aggression call that is structurally and aurally similar to it. Most of these vocalizations had a frequency modulation pattern wherein the frequency rapidly fell then rapidly rose, and these modulations were repeated from two to twenty or more times without a break. The number of times that the pattern was repeated varied greatly between the many different song components. Spectrograms of these song elements and the aggression calls looked like a zigzag pattern; as if a number of 'v' shaped figures had been run together. In some of these, the amplitude of each of the rising portions, or each of the falling portions was greatly reduced. The raspy sounding units were trills and it appears that some of them could have evolved from a buzzy element by omitting the rising portions of a buzzy element. The rates at which the frequency changed varied slightly between the different kinds of song components. Each was given in a frequency band that was different than that of other buzzy or raspy components. Many of these were usually given at the ends of highly stereotyped phrases and in each phrase they were preceded by a few simple elements, which varied between the different phrases.
Track Title: Aggression Calls and Similar Song Elements
The track entitled: Distinctive T 10 Song A 1 is one example of a phrase composed of buzzy sounding elements used by Type 10. It was heard occasionally, and was most often given towards the end of a song bout. The most frequently heard song component used by Type 10 in reactions to other males (as during matched counter-singing) was a raspy trill with a long, down swept element at the end rather than a single continuous element. The track designated as: Distinctive T 10 Song A 2 is the stereotyped phrase using this raspy unit. It is so distinctive that it might be possible to identify Type 10 crossbills by this song.
Track Title: Distinctive T 10 Song A 1
Track Title: Distinctive T 10 Song A 2
Type 2 and Type 3 have songs that may distinguish them as well. In the examples of distinctive songs below, the
songs are not buzzy sounding.
Track: Distinctive T 2 Song A – 1 is a distinctive Type 2 song that is a portion of the recording on: T 2 Song A 2. It appears to be one of their most frequently used songs. It was given briefly by two birds during December 2004 following an invasion in Humboldt County, California and was the only song that Type 2 used. It was also recorded in 2006; east of Crater Lake in Oregon. Track T 2 Song A 2 is a song bout that was recorded by Ron LeValley near Mendocino, north of San Francisco, California in 2003. This track is an abridged version of Ron's recording.
Track Title: Distinctive T 2 Song A - 1
Track Title: Distinctive T 2 Song A 2
Track: Distinctive T 3 Song A – 1 is a very distinctive unit and it was the song component most often heard from Type 3 during this study. From January to April 2005 it was the only Type 3 song heard. Track: T 3 Song A 2 is a short song bout and includes the song unit that is on: Distinctive T 3 Song A – 1.
Track Title: Distinctive T 3 Song A - 1
Track Title: Distinctive T 3 Song A 2
In some bouts of singing, Red Crossbills used only a single element, unit or phrase that was repeated in a series followed
by a brief silent pause, then this pattern was repeated, becoming a routine. A few birds used only the same simple
routine for days or even weeks. Males more typically used 3 or 4 song components, each repeated in a series and each
series was followed by a series of a different song component; then this sequence was repeated with the same order of
delivery. This rather simple routine characterizes relatively shorter song bouts (generally less than one minute).
Certain phrases (or short stereotyped songs) were not repeated in a series. Most of these phrases were composed of
high frequency elements that were short in duration. They seemed to be used like an introduction to a series of one
or another element or unit, which was usually followed by a series of a different component.
Many routines began with one of these phrases. Usually during longer bouts of singing (~1-5 minutes), following several repetitions of a routine, one of the song components that was used was replaced with a series of a different component of the repertoire. In this way the entire routine was changed over a moderately short period of time and subsequently modified further as the bout continued. Although these birds changed components many times during the bout they used only a small portion of the repertoire.
On rare occasions some males seemed to be more highly motivated. Song bouts were longer in duration (~5-10 minutes) and were much more varied. They tended to make changes to routines after only a few repetitions, and at times dropped the routine and began singing an entirely new one. During some of the most diverse bouts, birds seldom used routines. They often changed from a series of one component to a series of a component that was different than any used in a previous sequence. In most cases, Type 10 males utilized less than half of the song repertoire during the longest and most varied bouts.
Of these highly varied song bouts, those that were recorded later in the reproductive period gave the impression that they tracked or recapitulated the changes in the use of song components that seemed to have occurred over the course of the reproductive period. Each of these typically began with short, relatively simple components. Later in the bout many long buzzy and raspy elements, units or phrases were incorporated into the song and therefore the later portion was reminiscent of a bird that was counter-singing. Due to the inclusion of these long, buzzy elements the later parts of these bouts were more dramatic sounding than earlier portions.
Track: T 10 Long Song Bout A is a representative sample of a late season, long, and highly varied Type 10 song bout.
The track: T 2 Moderate Song Bout B is a bout recorded from an individual (designated as B) recorded 9 km east of Crater Lake, Oregon. It is moderately long in duration and this bird used a moderate proportion of the Type 2 repertoire during this bout. In these characteristics this bout is fairly typical of the song bouts voiced by many Red Crossbills.
Track Title: T 2 Song Like T 7 Flight Call
Track Title: T 2 Song A 2
Track Title: Distinctive T 2 Song A - 1
Track Title: T 3 Song Like T 2 Flight Call
Track Title: T 3 Song A 2
Track Title: Distinctive T 3 Song A - 1
Track Title: T 10 Song Like T 2 Flight Call
Track Title: Variations of a T 10 Song A 1
Track Title: Distinctive T 10 Song A 1
Track Title: Distinctive T 10 Song A 2
Adkisson, C. S. 1981. Geographic variation in vocalizations and evolution of North American Pine Grosbeaks. Condor
Benkman, C. W. 1993. Adaptation to single resources and the evolution of crossbill (Loxia) diversity. Ecological Monographs 63:305-325.
Benkman, C. W. 1999. The selection mosaic and diversifying coevolution between crossbills and Lodgepole Pine. American Naturalist 153: supplement 75–91.
Benkman, C. W. 2003. Divergent selection causes the adaptive radiation of crossbills. Evolution 57:1176–1181.
Benkman, C. W., and Miller, R. E. 1996. Morphological evolution in response to fluctuating selection. Evolution 50:2499–2504.
Benkman, C. W., Smith J. W., Keenan P. C., Parchman T. L. and Santisteban L. 2009. A new species of the Red Crossbill (Fringillidae: Loxia ) from Idaho. Condor 111:169-176.
Cramp, S. and Perrins, C. M. (editors). 1994. The birds of the Western Palearctic. 8. Oxford.
Edelaar, P. M., Robb, M., van Erde, K., Terpstra, K., Bijlsma, R. and Massen, E. 2004. Zijner meerdere soorten "Gewone" Kruisbek in Nederland? Limosa 77:31-38.
Harris, A. S. 1990. Picea sitchensis, in Silvics of North America: 1. Conifers (R. M. Burns and B. H. Honkala, tech. coords). Agriculture Handbook 654. U.S. Dept. Agric., Forest Service, Washington, D.C.
Groth, J. G. 1990. Cryptic species of nomadic birds in the Red Crossbill (Loxia curvirostra) complex of North America. PhD dissertation. University of California at Berkeley.
Groth, J. G. 1993a. Evolutionary differentiation in morophology, vocalizations, and allozymes among nomadic sibling species in the North American Red Crossbill (Loxia curvirostra) Complex. University of California Publications in Zoology, Berkeley. 127:1-143.
Groth, J. G. 1993b. Call matching and positive assortative mating in Red Crossbills. Auk 110:398-401.
Irwin, D. E. and Price, T. 1999. Sexual imprinting, learning and speciation. Heredity 82:347-354.
Irwin, K. 2010. A New and Cryptic Type of the Red Crossbill. Western Birds 41:10-25.
Keenan, P. C. and Benkman, C. W. 2008. Call imitation and call modification in Red Crossbills. Condor 110:93-101.
Mundinger, P. C. 1970. Vocal imitation and individual recognition of finch calls. Science 168:480-482.
Mundinger, P. C. 1979. Call learning in the Cardulinae: ethological and systematic implications. Systematic Zoology 28:270-283.
Nethersole-Thompson, D. 1975. Pine Crossbills. T. & A. D. Poyser, Berkhamsted.
Newton, I. 1972. Finches. Collins, London.
Orr, M. R., and Smith, T. B. 1998. Ecology and speciation. Trends in Ecology and Evolution 13: 502-506.
Parchman, T. L., Benkman, C. W. and Britch, S. C. 2006. Patterns of genetic variation in the adaptive radiation of New World crossbills (Aves: Loxia). Molecular Ecology 15:1873-1887.
Robb, M. S. 2000. Introduction to vocalizations of crossbills in north-western Europe. Dutch Birding 22:61-107.
Rundle, H. D., and Nosil, P. 2005. Ecological speciation. Ecology Letters 8: 336-352.
Sewall, K. B. 2008. Age dependant plasticity of call production and responsiveness in red crossbills (Loxia curvirostra): implications for social intermixing and evolution. Ph.D. thesis, University of California at Davis.
Smith, J. W., Benkman, C. W. and Coffey, K. 1999. The use and misuse of public information by foraging Red Crossbills. Behavioral Ecology 10:54-62.
Smith, J. W. and Benkman, C. W. 2007. A coevolutionary arms race causes ecological speciation in crossbills. American Naturalist 169:455-465 (and appendix online).
Snowberg, L. K. and Benkman, C. W. 2007. The role of marker traits in the assortative mating within red crossbills, Loxia curvirostra complex. Journal of Evolutionary Biology 20:1924-1932.
Summers, R. W., Dawson, R. J. and Phillips, R. E. 2007. Assortative mating and patterns of inheritance indicate that the three crossbill taxa in Scotland are species. Journal of Avian Biology 38:153-162.
|Bill depth at nares (mm)||Bill length from nostril (mm)||Tarsus||Wing chord||Mass (g)|
|Bill depth at nares (mm)||Bill length from nostril (mm)||Tarsus||Wing chord||Mass (g)|
|Bill depth at nares (mm)||Bill length from nostril (mm)||Tarsus||Wing chord||Mass (g)|