This article was originally published in Oregon Birds 16(3):218-221. I continued to capture birds after this was published, increasing the data base with no substantial change in the conclusions.

Green-backed Selasphorus Hummingbirds
in Clatsop County, Oregon

Mike Patterson,

During the hummingbird migration in the spring of 1988 I caught two anomalous hummingbirds which I tentatively identified as displaced Allen's Hummingbirds (Selaphorus sasin). In reporting these (Patterson, 1988), I mentioned several things that I intended to study further. It is the purpose of this report to describe the most recent data I have collected regarding the status of Selasphorus hummingbirds on the North Coast of Oregon and its relation to the male and female hummingbirds I captured in 1988.

The 19 88 male had a large amount of green in the back. Most of the upper back was green and the center of the lower back was green. Measurements for the tail and rectrices fell on the borderline between Allen's and Rufous Hummingbirds (S. rufus) as defined by the key I was using (Stiles, 1972). The second retrices showed very slight emargination on the inner webs. The 1988 female also had borderline measurements for tail and rectrices.

In an effort to confirm the identifications made of the m ale and female hummingbirds caught 1988, I sent their measurements to Dr. Gary Stiles in Costa Rica and Dr. Robert Bleisweiss in Massachusetts. Both agreed that the measurements were small, but stopped short of confirming the identification of the two birds as Allen's Hummingbirds. They made some very good suggestions regarding things to look for in the next season. These included attempting to age birds as to whether they are older or younger than 2 years by looking at marks on their culmens (Ortiz-Cr espo, 1972) and making a comparison series of measurements to test how closely my measurements conform to those made by Dr. Stiles when he produced his key (1972). I also wanted to make some quantitative assessment of green-backed males with the intent of correlating this to age, mensural characters and time of arrival.

Green-backed Birds

To quantify the green on male birds I divided the back into four equal regions and assessed the amount of green assigning a number corresponding to quantity of green found (Table 1,Fig. 1). The four numbers could then be averaged together to provide a relative index of green on the backs of males.

The first Rufous Hummingbird arrived at my feeder on February 28 (the same date as in 1988). I caught birds through March and April, 55 birds in all; 37 males, 18 females. All 1989 birds had measurements that would identify them as Rufous Hummingbirds, falling well within the boundaries defined by Stiles ' key. Of the 37 males captured, all but 6 showed some green on the back. The average amount of green fell at 0.85 on the scale or about 20% green (Fig. 1-a). Ten males were caught with at least 30% green on their backs (Fig. 1-b) and one had over 50% green (Fig. 1-c). This last male had an upper back that was almost entirely green and green in the center of its lower back. One-third of all the male birds captured had enough green in the back to probably be observable in the field.

Fernando Ortiz-Crespo (1972) found that immature birds showed fine diagonal corrugation on their bills. These corrugation remain at least partially on most first-year adults. I looked for these corrugations in the hope of relating age and amount of green on the back. Though bill markings indicative of first year birds were found, the number (8 males and 5 females) was too small for reliable statistical analysis.

The re did appear to be a slight trend toward very green-backed males to appear later in the season. Taking the median for total number of males (March 30) as the mid-point and all males with above average green (17 birds), 41% appeared before the median and 59% after the median. It is difficult to say whether this is truly significant. Perhaps analysis of larger samples will make this more clear.

It is clear that male Rufous Hummingbirds with green backs are the rule rather than the exception at l east on the North Coast of Oregon. Most of these males have only very small amounts of green on them and this would only be discernible at fairly close range. A significant number do have what could be considered large amounts of green in the back and t his tends to be concentrated at the upper center of the back. Other measurements on these birds do not appear to be appreciably different from red-backed birds. From this I would suggest that a green-backed male Selasphorus hummingbird outside the expected range of Allen's Hummingbirds is almost certainly a green-backed Rufous. The only way to conclusively prove otherwise would be to somehow get the bird in hand, measure the tail and rectrices and examine the 2nd rectrix for emarginations as was done with the two birds caught in 1988.

Comparison of Measurements with Key

In using the key prepared by G. F. Stiles (1972) for separating Selasphorus hummingbirds, one has to assume that two observers are comparable in their measuring techniques. In an effort to test the comparability of my measurements with this key, I carefully measured all birds captured using vernier calipers to the nearest tenth of a millimeter. I also collected 1st, 2nd and 5th rectrices from the right side of the tail of all birds caught.

A comparison of measurement averages (Table 2) shows that mine are larger in almost every case. Comparison of standard deviations show that my measurements are as consistent as his. However, my measurements average 4.6% larger than those of Stiles overall and rectrix widths averaged 11% larger. The most probable explanation comes from the fact that Stiles made measurements for his key from study skins and all of my measurements were on living spec imens. Study skins tend to shrink up some. He measured rectrix width on the bird. I measured rectrix width on plucked feathers flattened out on a card.

What this suggests to me is that, if my measurements tend to be larger, than the anomalous male and female I captured in 1988 would not fit well into Stiles' key without adjusting for the difference between measuring techniques. This adjustment would tend to support their identification as a male and female Allen's Hummingbirds.

I would stress here that I am not suggesting the that the data presented by Dr. Stiles is any less correct than my own. Differences are almost certainly attributable to differences in technique and only serve to stress the necessity for using caution when u sing any key based upon measurement data.

Anomalous Birds

The amount of green on the back of the March 10, 1988 male was greater then that of any male bird caught in 1989, but it did have red on the margins of its lower back back and w ould probably rated about a 3.5 on my scale. The male Hummingbird obtained from Philomath in 1983 and identified as an Allen's (Patterson, 1986) is a 3.75 with more red in the upper back and less in the lower back. The photograph of an Allen's Hummingbi rd in Tyrrell's Hummingbirds (1985) shows a great deal of red in the lower back and rates a 3.0. This suggests that back color in Allen's Hummingbirds is highly variable and there are overlaps between male Allen's and Rufous Hummingbirds. An analysis of back color in a large sample of living Allen's Hummingbirds to determine the exact degree of overlap would would be of interest.

The only reliable means of separating females is by measuring tail length and rectrix width. Considering the apparent overlap of back color for males, tail measurements and 2nd rectrix emargination, may be the only reliable way to separate green-backed males, as well. Both of the birds caught in 1988 were small, but borderline according to Stiles' key. Since my measurements average larger, appropriate adjustments would place both of these birds well within the measurements for Allen's Hummingbirds.

Obviously, the only way to have been absolutely certain of the identity of these birds would have been to collect them and compare them to other specimens. Needless to say I did not collect them, but based upon their measurements and the mensural data I have collected on other North Coast hummingbirds, I am very confident that they were both Allen's Hummingbirds. The fact that I collected a larger sample in 1989 and did not catch any small birds suggests that Allen's Hummingbirds are accidental this far north of their expected range and that the majority of green-backed males seen this far north are most probably Rufous Hummingbirds.


Table 1: quantitative scale for determining amount of green in male 
Selasphorus hummingbirds and percent of males found in each category.

color index                            % males
0  -  no green                          16.7%
1  -  1 to 25 % green                   52.7%
2  -  26 to 50 % green                  27.8%
3  -  51 to 75 % green                   2.8%
4  -  76 to 99 % green                   0
5  -  all green                          0

Table 2: Comparison of average measurements of Rufous Hummingbirds 
made by Stiles and Patterson 

                Stiles' measurements               Patterson's measurements
                M      sd     F       sd              M     sd      F    sd
n               30            30                      37            22
culmen         16.07  .60    17.63   .64             16.8  .54     18.1  .58
wing           40.32  .87    44.40   .81             40.5  .82     44.0  .80
tail           27.36  .91    25.92  1.03             28.2  .97     27.0  .90
1st rectrix                   7.96   .29              9.3  .36      8.9  .39
5th rectrix                   3.34   .31              2.6  .18      3.7  .25