A Critique of Wang Yong and Finch's Field-Identifications of Willow Flycatcher Subspecies in New Mexico [with Response]

8 Pages·1999·3.9 MB·English

by  John P Hubbard

Wilson Bull., I 1 1(4), 1999. pp. 585-588
Commentary
A CRITIQUE OF WANG YONG AND FINCH’S
FIEFD-IDENTIFICATIONS OF WIFFOW FFYCATCHER
SUBSPECIES IN NEW MEXICO
John Hubbard'
P.
In a recent paper in the Wilson Bulletin, reliable for this purpose in the Willow Fly-
Wang Yong and Finch (1997; henceforth catcher.
Y&F) reported that they subspecifically iden- As my earlier comments suggest, I do not
Y&F
tified 83 of 84 Willow Flycatchers {Ernpidon- believe used the approach described
ax traillii) captured, banded, and released in above in their attempts to identify subspecies
central New Mexico in spring and autumn in the Willow Flycatcher. In other words, they
1994 and 1995. Given the nature of these sub- did not take synoptic series of study skins into
species and the means by which Y&F appar- the field, against which the birds they captured
ently identified them, I am extremely doubtful were compared to determine subspecific iden-
about the reliability of their determinations tities. However, I cannot be 100% certain
and thus the validity of these as scientific data. about this because the methods section in their
The fact is that identifying these taxa is quite paper is so incomplete and otherwise deficient
difficult, even for trained taxonomists working one can only guess at many aspects of their
in the laboratory under the best protocols and approach. Nonetheless, it seems logical that if
conditions. This difficulty stems from a num- they had used skins as the basis for their iden-
ber of factors, the major one being the per- tifications, they would have said so. Given this
vasive subtlety of the plumage-color charac- assumption, if they did not use skins, how did
Nteorts sbuyrpwrihsiicnhglyt,hetsheesseubdsifpfeecrieenscemsaiarnelydifdfiifcfuerl.t tshpeeyciegso? aObonutthiidsentmiaftytienrg tYhe&iFr biarrdes atto sbuebs-t
to describe in words, which is exacerbated by vague, providing a few clues but no definitive
the fact that none of the available classifica- explanations of their identification methodol-
tion systems accurately portrays the range of ogy. For example, we are told that they
plumage coloration observed in this flycatcher “. . . adopted the four-subspecies classification
(e.g.. Browning 1993). This means that this system of Hubbard (1987) and Unitt (1987),
species’ plumage-color characters are best ob- in which “subspecies identity ... is based [in
served in specimens (i.e., study or flat skins), part] on . . . coloration of the head [= crown]
which also provide the best avenue for iden- and neck [= forenape] and its contrast with
tifying subspecies. To do this, one must first the back, and the contrast between the breast-
band and the throat (see Phillips 1948, Hub-
assemble series of skins representing all rel-
evant taxa, as well as such important subcat- bard 1987, Unitt 1987, BrowYn&iFng 1993).”
egories as age classes (e.g., adult vs immature) Based on this, I assume that chose lit-
and seasonal groupings (e.g., spring vs au- erature descriptions (as opposed to specimen
tumn). Then one sorts “unknowns” (which comparisons) as the basis for their identifica-
could include live birds) into subcategories tion of Willow Flycatcher subspecies. In ad-
and compares them to the taxa therein, which dition, 1 also suspect they converted these de-
should produce at least tentative subspecific scriptions into the color values of Smithe
identifications. In fact, this is the standard lab- (1975), as this is the system they used to clas-
oratory approach for identifying color-based sify coloration in birds captured in the field.
subspecies, and it is the only means proven Beyond this, one could also speculate on such
matters as (a) how converted values were ac-
tually used to identify birds, e.g., whether in
' Route 5, Box 431, Espanola, New Mexico 87532. a dichotomous key, probability table, or other
585
586 THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999
framework; or (b) what Y&F’s perceptions example, in the latest revision of the species.
were of color characters in various races, giv- Browning (1993) could only assemble 270
—
en that no such descriptions were offered by specimens of breeding season adults includ-
them. However, I see no purpose in further ing fewer than 20 of the endangered subspe-
speculation concerning these or other aspects cies E. t. extirnus of the Southwest. As a con-
of their methodology. This is because if they sequence, it is not surprising that he ques-
did base their identifications on the literature tioned boundaries between four of the five
rather than specimens, 1 believe the process subspecies recognized in his paper. Even
beca—me so flawed that the details are irrele- when populational characteristics are better
vant like rearranging deck chairs on the known, opinions may differ as regards their
sinking Titantic! taxonomic treatment. Thus, Browning (1993)
The message here is that the literature is no recognized two subspecies (i.e., E. t. trailHi
substitute for specimen comparisons for any- and E. t. campestris) as breeding in the region
one attempting to identify Willow Flycatcher east of the Rocky Mountains, whereas Unitt
subspecies, at least if attaining the most reli- (1987) merged the latter with the nominate
able scientific data is the goal. Furthermore, form. Differences in opinion also exist on a
given logistical and other problems, 1 doubt broader scale, such as concerning the overall
even specimen comparisons would consistent- number of subspecies recognizable in the Wil-
ly yield reliable identifications of live birds low Flycatcher. For example, some taxono-
under field conditions. Not only would it be mists maintain that none should be recognized
unwieldy to take and use museum skins in the (e.g., Mayr and Short 1970, Traylor 1979),
field, but setting up and maintaining constant while others accept four to six as valid (e.g.,
conditions (e.g., lighting) would also be dif- Phillips 1948, Aldrich 1951, Wetmore 1972,
ficult. In addition, except for recaptures, only Oberholser 1974, Unitt 1987, Browning
one opportunity would be available to identify 1993). Thus, although specimen comparisons
each live bird in the field. This means that one provide our only reliable means for identify-
could not reassess identifications at a later ing subspecies in this flycatcher, this approach
time, which is both frequent and necessary must be used with the clear recognition that it
when studying specimens in the laboratory. In is just the first step in this very difficult en-
this regard, photographs and certainly color deavor.
readings (e.g., from Smithe 1975) would not Incidentally, the above differences in taxo-
be adequate for such reexaminations because nomic opinion present a problem for those
these do not exactly duplicate colors observed that rely largely or entirely on the literature
in the birds or specimens themselves. Given for their knowledge of geographic variation in
these considerations, I believe that identifying this species. That is, how does one choose
subspecies in the Willow Flycatcher is best which authorities to follow and thus which
done in the laboratory, using study skins ex- viewpoints to accept on this subject? Among
amined under proper protocols and procedures others, one way around this would be to ad-
by people trained in the process. In other here strictly to a single point of view, such as
words, this is a task that should be left to an the recent revision of this flycatcher by
alpha-taxonomic approach, which is appropri- Browning (1993). However, Y&F chose not
ate when one considers that subspecies arose do this, instead electing to cobble their con-
and largely remain as products of that realm. cept of variation from a variety of sources
Even when approached as outlined above, (e.g., Phillips 1948, Hubbard 1987, Unitt
the reality is that not every specimen or even 1987, Browning 1993). Given the lack of con-
population of this flycatcher can be reliably sensus among these sources, this was a ques-
assigned to subspecies. Intergradation and tionable decision. In fact, it would be a chal-
overlap occur in all characters that distinguish lenge even for people with firsthand experi-
these taxa, so birds exhibiting such character- ence with geographic variation in this species,
istics may be un- or misidentified as a result. as seen from the variety of opinions cited
In addition, characteristics in some popula- above. As a consequence, it is not surprising
tions remain poorly known, mainly because of that I would quibble with Y&F’s choices, in-
the paucity of specimens from these areas. For cluding that of which authorities to follow.
Huhhard COMMENTARY 587
•
For example, as indicated earlier, they cited this form, even though the the middle Rio
my unpublished paper (Hubbard 1987) as a Grande Valley lies some 200 miles west of the
basis for the “four-subspecies classification nearest specimen localities in New Mexico.
system” adopted in their study. However, that Finally is the race E. t. adastus, which breeds
so-called system was actually a cobbling job widely in the interior U.S. north of the .south-
itself, my aim being to summarize color char- western states, through which it passes in both
acters of various subspecies from the treat- spring and autumn. In New Mexico, it com-
ments of Phillips (1948), Aldrich (1951), Wet- prised 25.6% of the specimens reported by
more (1972), and Oberholser (1974). As such, Hubbard (1987), compared to 10.8% in Y&F’s
it was not meant either to provide definitive sample.
descriptions of these subspecies or to recom- As for the fourth subspecies {E. t. extimus),
mend which should be recognized as valid. Y&F identified 34 (41.1%) of their birds as
For it to have been otherwise used by Y&F this form, compared to the 48.8% from
may seem flattering, but it certainly was not a throughout New Mexico by Hubbard (1987).
sound decision from a taxonomic viewpoint. Thus, on the face of it, their findings would
Given the flawed nature of their approach, seem not to differ significantly from what is
it is no surprise that Y&F’s findings on Wil- known from specimens of this taxon. How-
low Flycatcher subspecies would also be open ever, the number of questionable literature re-
to question. For example, when compared cords of this subspecies suggests it may be
with what is known from specimens (e.g., more subject to misidentification than certain
Hubbard 1987), significant differences emerge other forms, such E. t. brewsteri and E. t.
on the New Mexico status of three of the four traillii (both sensu lato). Birds that might be
taxa recognized in that study. (In light of the mistaken for E. t. extimus could include sun-
relative scientific standing of the two sources, bleached or worn individuals of other races,
I would obviously accept the specimen ver- as well as pale variants of E. t. adastus, in-
sion over that of Y&F in every case.) The tergrades between the latter and E. t. extimus,
most significant difference occurs in the sub- and carelessly-examined E. t. campestris. If so
species E. brewsteri (sensu stricto), which misidentified, such instances could help ex-
t.
breeds along the Pacific slope of North Amer- plain records of E. t. extimus from areas out-
ica. Although occurring regularly in migration side its known breeding range, such as the
eastward to Arizona (Monson and Phillips northern two-thirds of Colorado (Bailey and
1981), this form has rarely been collected east Niedrach 1965) and Texas east of the Trans-
and north of that state, e.g., in Utah (Behle Pecos region (Oberholser 1974). As for New
1985), Colorado (Bailey and Niedrach 1965), Mexico, 1 am dubious of E. t. extimus records
Oklahoma (Sutton 1967), and Texas (Ober- from the eastern plains, such as two speci-
holser 1967). Hard data from New Mexico mens reported in Hubbard (1987) from Roo-
clearly conform to this pattern, with only two sevelt County. In addition, 1 have definitely
(4.7%) of the 43 specimens so attributed in reidentified two of the purported E. t. extimus
Hubbard (1987) and even these were some- from that report, one from San Juan County
what equivocal. By contrast, Y&F identified (= E. t. adastus > extimus) and another from
>
33 (39.8%) of their 83 birds as E. t. brewsteri, Socorro County (= E. t. extimus adastus).
which is about 8.5 times more frequent than Of course, as mentioned earlier, we do not
reported by Hubbard. Another notable depar- have the luxury of reexamining E. t. e.xtimus
ture involves the subspecies E. t. traillii (in (or other subspecies) reported by Y&F, so
which Y&F include E. t. campestris), which their identifications cannot be reassessed in
breeds from the Great Plains to the northeast- light of potential sources of misidentification.
ern Atlantic Coast. In the Southwest, E. t. Given this and their flawed methodology, 1 see
traillii/campestris occurs regularly in the no reason to regard their findings on this form
plains of eastern Colorado (Bailey and Nied- as any more acceptable than those on the other
rach 1965) and New Mexico (Hubbard 1987), races reported in their paper. As a final point,
but it has not been collected as far west as Y&F make no mention of the differences be-
Arizona (Monson and Phillips 1981). Yet tween their findings on the various subspecies
Y&F reported that 8.4% of their birds were and the specimen record as discussed above.
588 THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999
While the need for this would not have been LITERATURE CITED
obvious as regards E. extimus and perhaps
t.
even E. t. adastus, this could hardly have been Aldrich, J. W. 1951. A review of the races of Traill’s
the case with E. t. traillii and especially E. t. Elycatcher. Wilson Bull. 63:192-197.
brewsteri. Bailey, A. M. and R. J. Niedrach. 1965. Birds of
To summarize, geographic variation in the Colorado, vol. 2. Denver Museum of Natural His-
tory, Denver, Colorado.
Willow Flycatcher mainly involves subtle dif-
Behle, W. C. 1985. Utah birds: geographic variation
ferences in plumage coloration, concerning
and systematics. Utah Mus. Nat. Hist. Occ. Pub.
which taxonomists disagree in terms of the 5:1-147.
number of subspecies that should be recog- Browning, M. R. 1993. Comments on the taxonomy
nized. Anyone contemplating identifying of Empidonax traillii West. Birds 24:241-257.
these subspecies should do so with these ca- Hubbard, J. P. 1987. The status of the Willow Fly-
veats in mind, as well as by approaching the catcher in New Mexico. Unpublished report. New
process through the use of specimen compar- Mexico Dept, of Game and Fish, Santa Fe.
— Mayr, E. and L. L. Short, Jr. 1970. Species taxa of
isons preferably in the laboratory under con-
North American birds. Nuttall Ornithithological
trolled conditions and with proper training in
Club, Cambridge, Massachusetts.
alpha-taxonomic procedures. Given that Monson, G. and a. R. Phillips. 1981. Annotated
Y&F’s approach appears to have been other- checklist ofthe birds ofArizona. Univ. ofArizona
wise, I submit that their field identification of Press, Tucson.
these subspecies cannot be regarded as a bona Oberholser, H. C. 1974. The birdlife of Texas, vol.
fide assessment of this parameter in the birds 2. Univ. of Texas Press, Austin.
they processed in New Mexico in 1994 and Phillips, A. R. 1948. Geographic variation in Etnpi-
1995. Furthermore, for those that would use donax traillii. Auk 65:507-514.
their subspecific findings, I urge them to do Phillips, A. R., J. Marshall, and G. Monson. 1964.
The birds of Arizona. Univ. Arizona Press, Tuc-
so with extreme caution to say the least. Be-
son.
yond this, I would like to state that as an al- Smithe, F. B. 1975. Naturalist’s color guide. American
pha-taxonomist, I am dismayed that a study Museum of Natural History, New York.
with such a flawed approach to subspecies Sutton, G. M. 1967. Oklahoma birds. Univ. of
identification could make its way into print in Oklahoma Press, Norman.
a major ornithological journal. To wit, orni- Traylor, M. A., JR. 1979. Check-list of birds of the
thology has come to rely almost entirely on world, vol. 8. Harvard Univ., Cambridge, Massa-
non-specimen data for monitoring the distri- chusetts.
bution and status of birds on this planet. While Unitt, P. 1987. Empidonax traillii extimus'. an endan-
gered subspecies. West. Birds 18:137-162.
not necessarily a bad thing, sometimes we
Yong, W. and D. M. Finch. 1997. Migration of the
may fail to recognize the very real limitations Willow Flycatcher along the Middle Rio Grande.
of such data. No better example of this exists Wilson Bull. 109:253-268.
than as regards the identification of difficult Wetmore, a. 1972. The birds of the Republic of Pan-
taxa, of which subspecies in Empidonax trail- ama, part 3. Smithsonian Misc. Coll. 150(3):1-
lii provide a perfect case in point. 631.
Vonf; and Finch • RESPONSE 589
Wilson Bull., 111(4), 1999, pp. 589-592
RESPONSE
Wang Yong' and Deborah M. Finch* -
Hubbard (1999) criticizes our paper Migra- species skins as advocated by Hubbard to
tion ofthe Willow Flycatcher along the middle identify Willow Flycatcher subspecies in the
Rio Grande (Yong and Finch 1997), where we field, we relied on descriptions and records of
reported aspects of stopover ecology of the coloration and morphology published in the
species including timing, abundance, fat available literature by taxonomists. Contrary
stores, stopover length, and habitat use. Hub- to what Hubbard speculates, we did not con-
bard questions our identification of subspecies vert color descriptions into Smithe’s (1975)
of the Willow Flycatcher {Empidonax traillii) color code values. We based our identification
and the methods we used to identify them. He of back plumage color on the most recent re-
also attempts to evaluate the accuracy of our search by Unitt (1987) and Browning (1993).
results of subspecies composition by compar- Using Smithe’s color codes to describe back
ing them with data from other researchers. We plumage, Unitt (1987) writes: “In brewsteri
welcome and applaud this scrutiny in the hope the green is in the direction of olive green
that this interchange will stimulate greater in- (color 48), in adastus in the direction of
terest, research, and capability to distinguish greenish olive (color 49), and in extimus and
the phenotypic characteristics of subspecies of traillii in the direction of grayish olive (color
the Willow Flycatcher. Given that the south- 43). That is, brewsteri is a dark brownish ol-
western race {E. t. extimus) of the Willow Fly- ive, adastus a dark grayish green, and extimus
catcher is federally listed as Endangered, re- and traillii a pale grayish green. .’’ Browning
.
liable methods for identifying this subspecies (1993) suggested that Smithe’s color system
need to be developed to more effectively con- is problematical because the color swatches
serve and recover its populations. generally are not identical matches for actual
We are aware that the subspecific taxonomy colors. Hence, he used Munsell Color Charts
of the Willow Flycatcher is inconsistent (1990) to describe the crown and back con-
among taxonomists as are the techniques to trast for his specimens. During our fieldwork,
identify subspecies. Consequently, reliable we consulted both Unitt’s (1987) color codes
identification of subspecies is difficult, espe- for subspecies’ back color and Browning’s
cially in field situations. We acknowledge that color contrast scores between crown and back.
issues of taxonomic status, population distri- Although Hubbard suggests that live speci-
butions, and identification methods of subspe- mens have some disadvantages, we counter
cies of the Willow Flycatcher should be ex- that the plumage coloration of live birds is
plored further. However, Hubbard’s criticisms more likely to be true to type than skin spec-
of our paper are generally based on erroneous imen plumage that may have faded. If our hy-
information as well as incorrect assumptions pothesis that the coloration of fresh plumage
about our methods, and they do not alter our differs from that of faded plumage is correct,
conclusions about Willow Flycatcher stopover then data collected from live specimens may
ecology at the species level. be more reliable, or at least not less reliable,
Hubbard’s first criticism focuses on the than results obtained from study skins. Birds
methods we used for identifying the subspe-
occasionally called or sang in our study after
cies. Rather than using an assemblage of sub- being released. Information about song and
call characteristics were also recorded when
' USDA Forest Service, Rocky Mountain ForestNReM- possible. Such data are available from living
search Station, 2205 Columbia, SE, Albuquerque, flycatchers but not from skins. Sedgewick’s
87106.
^ Corresponding author; E-mail: Finch_Deborah_M/ (pers. comm.) preliminary analyses of Willow
Flycatcher song and call signatures collected
[email protected]
590 THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999
in different regions suggest that E. t. extimus achievable.” Although taxonomists disagree
song structure can be distinguished from that in their interpretations of within-species vari-
of its northern conspecifics and we used this ation and subspecies recognition, there is un-
kind of data to aid identification also. mistakable agreement about use of a four sub-
We
did not rely solely on coloration for species classification among recent research
subspecies identification, contrary to Hub- papers (Hubbard 1987, Unitt 1987, Browning
bard’s second assumption. Unitt (1987) sug- 1993). Hubbard (1987) clearly advocates ac-
gested that wing formula (relative length of ceptance of the four subspecies classification
primary feather length) can be used to assist in his report by stating that: “Given the degree
subspecies identification. Of the 305 speci- of agreement among recent workers, I believe
mens that Unitt (1987) examined, wing for- the most prudent course is to accept all of the
mula distinguished 93% of the E. t. extimus above subspecies [i.e., E. t. extimu—s, brewsteri,
and E. t. traillii, 88% of the E. t. adastus and and adastus^ and traillii as valid at least un-
E. t. traillii, and 89% of the E. t. brewsteri til more definitive studies are available.” Al-
and E. t. traillii. Browning (1993) also applied though in his commentary Hubbard declares
wing formula to assess variation in subspecific his own report to be a “cobbling job”, its
characteristics, and his results also demon- quality is deemed sound by other authorities.
strated that wing formula may be useful for Indeed, it has been widely distributed and cit-
distinguishing some subspecies although his ed both unofficially and officially by the En-
sample size was smaller than Unitt’s. Hubbard dangered Species Programs of U.S. Fish and
himself (1987) noted that E. t. brewsteri was Wildlife Service regions, by state Game and
smaller than other described forms. In the Fish Departments, and by other agencies and
field, we relied partly on non-overlapping ex- ornithologists in the western United States, es-
treme wing measurements to assist in the pecially in the Southwest. Given Hubbard’s
identification of this subspecies. In addition, background as a competent taxonomist in
we measured and recorded more than 30 var- New Mexico and as an officer of the state en-
iables from each individual. Following Unitt dangered species branch, his paper is judged
(1987), we used wing formula to aid in iden- as an authoritative source on the species. For
tifying subspecies. example, in the process use.d for listing the
Thirdly, Hubbard (1999) comments that southwestern Willow Flycatcher as a federally
“even when characteristics of populations are endangered subspecies, Hubbard’s paper was
better known, opinions may differ as regards one of the most heavily cited reports by the
their taxonomic treatment” because of limited U.S. Fish and Wildlife Service (1995).
sample sizes, interbreeding among populations, Unitt (1987) also states that the four races
and differences in taxonomists’ methods, of E. traillii are valid and may be distin-
views, and findings. Although we agree that guished from each other by “color, wing for-
taxonomists have been inconsistent in their mula, or both”. Browning (1993) further sep-
treatment of subspecific taxonomy, we consid- arated subspecies E. t. traillii into two popu-
er this to be an incentive for finding areas of lations: E. t. campestris of the Great Plains
common ground among researchers, rather and Great Lakes regions, and E. t. traillii to
than a justification for concluding that reliable the southeast of E. campestris. We recently
t.
identification of subspecies is impossible. became aware, that Unitt has conducted fur-
Hubbard states that we should have strictly ther research on the same specimens and may
adhered to a single view of subspecies tax- soon be updating his taxonomic treatment (P.
onomy. We followed a single view of subspe- Unitt, pers. com. through J. E. Cartron). These
cies treatment, but we did not credit this single different authors describe subspecies distri-
view to a single researcher. We made it clear butions that are very similar although popu-
that we adopted the “four subspecies classi- lation boundaries are not exactly the same.
fication system of Hubbard (1987) and Unitt U.S. Fish and Wildlife Service relied partly
(1987).” We warned readers in our Methods on these studies to conclude that listing the
section that: “Given morphological overlap southwestern Willow Flycatcher as an endan-
and hybridization among subspecies, complete gered subspecies was appropriate.
accuracy in identifying subspecies is not Fourthly, Hubbard evaluates our results by
Y(>nf> and Finch • RESPONSE 591
comparing our subspecies composition data species. Because E. t. extimus is endangered,
with subspecies data from his own and other U.S. Fish and Wildlife permits for collecting
reports and sources. While such comparisons voucher specimens during migration are not
may be valid for the purpose of exploring po- issued in the Southwest, eliminating the pos-
tential sources of viuriation, the conclusions sibility of having an alpha-taxonomist identify
that Hubbard draws are incorrect because of locally caught specimens to subspecies for the
spatial and temporal differences among stud- purpose of setting standards. Because most
ies. Species, subspecies, and population com- current research studies and conservation ef-
position of migratory birds captured at spe- forts pertaining to the Willow Flycatcher have
cific stopover sites in fall or spring can dra- focused on its breeding grounds, the impor-
matically differ from what is observed at the tance of our research centers on when, where,
same location during the breeding season at and how migration stopover sites in riparian
the same location or from other locations dur- woodlands along the middle Rio Grande are
ing migration. For example, the overall spe- used for resting and fat depositions by the spe-
cies composition we detected indicated that cies. Without understanding the migration
the majority of individuals captured were not strategy of the species and without justifying
local breeders and many did not even breed efforts to conserve the stopover habitat that
in New Mexico (Finch and Yong 1999). While the species uses, the Willow Flycatcher’s fate
we used a standardized, systematic procedure in the Southwest will be jeopardized regard-
to sample throughout the entire migration sea- less of how perfect or imperfect our ability in
sons of spring and fall, 1994 and 1995, other identifying subspecies is.
studies that Hubbard (1999) cites and com- Throughout ornithological history, subspe-
pares to ours were not conducted during mi- cies classification and identification have tra-
gration seasons and/or did not use standard- ditionally been a “problematic” area, partic-
ized procedures. In addition, source studies ularly within the genus Empidonax. Uncer-
cited by Hubbard are heterogeneous in rela- tainties about subspecies or even species sta-
tion to study goals, year of study, number of tus do not negate the value of our migration
years, geographical location, sampling design, research or refute our results about Willow
sampling season, and quality of data, leading Flycatcher stopover ecology or intraspecific
We
to uncontrolled and unknown factors that in- variation in migration patterns. assert that
validate comparisons with our data set. Our increased knowledge of the stopover behavior
data are restricted to two sites during two and energetic condition of the Willow Fly-
years in the middle Rio Grande valley of New catcher is important for understanding the bi-
Mexico, and thus are only truly comparable to ology of the species as a whole and that in-
other data from the same vicinity, year, and formation about within-species variation is
sampling design. Given that different studies, valuable in conserving the endangered south-
especially earlier ones, used controversial cri- western subspecies.
teria for classifying and counting their speci- Our paper and Hubbard’s (1999) critique
mens, Hubbard’s argument that our results are have opened up the opportunity to develop
inaccurate because they are not completely and expand discussion and evaluation of the
consistent with other studies that, when com- different subspecies, the subspecies concept as
pared, also yielded dissimilar results is circu- a whole, and whether subspecies should be
lar. In our manuscript, we did not make such recognized for the Willow Flycatcher given
comparisons for at least two reasons: (1) our the disagreement about their identification and
We
research focus was on the stopover biology of the difficulty in identifying birds in hand.
the species, not on the taxonomic status of the invite and challenge others to contribute ideas
subspecies, and (2) other data sources were and knowledge to this controversy in the hope
not homogeneous or similar enough to draw that new or better techniques for identifying
willow flycatcher subspecies may result. Such
comparisons.
Our data and conclusions about the fly- discussion or results would certify beyond a
catcher’s stopover ecology are not dependent doubt the worthwhile contribution of our pa-
on the validity or accuracy of its subspecies per. Subjecting any paper to a critical com-
status or on the methods used to identify sub- mentary, however, automatically attracts the
592 THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999
notice of additional readers. We are pleased catcher in New Mexico. New Mexico Dept, of
with the extra attention in the hope that further Game and Fish, Endangered Species Program,
research, understanding, and conservation ef- Santa Fe.
forts will be directed toward the endangered Hubbard, J. P. 1999. A critique of Wang Yong and
Finch’s Field-identification of Willow Flycatcher
southwestern Willow Flycatcher and its dis- subspecies in New Mexico. Wilson Bull. Ill:
appearing habitat. 585-588.
Munsell Color Charts. 1990. Munsell Color Com-
ACKNOWLEDGMENT
pany, Inc, Baltimore, Maryland.
We thank R. Banks for his review ofthe manuscript. Smithe, F. B. 1975. Naturalist’s color guide. American
Museum of Natural History, New York.
LITERATURE CITED Unitt, P. 1987. Empidonax traillii extiinu.s: an endan-
gered subspecies. West. Birds 18:137-162.
Browning, M. R. 1993. Comments on the taxonomy U.S. Fish and Wildlife Service. 1995. Final mle de-
of Empidonax traillii (Willow Elycatcher). West. termining endangered status for the Southwestern
Birds 24:241-257. Willow Flycatcher. Federal Register 60:10694—
Einch, D. M. and W. Yong. 1999. Landbird migration 10715.
in riparian habitats of the middle Rio Grande. Yong, W. and D. M. Finch. 1997. Migration of the
Stud. Avian Biol. In press. Willow Flycatcher along the middle Rio Grande.
Hubbard, J. R 1987. The status of the Willow Fly- Wilson Bulletin 109:253-268.