Brief Summary: New advances are presented in the field identification of Long-billed and Short-billed Dowitchers in worn alternate and basic plumages. Emphasis is placed on scapular and covert feather patterns and structural features in conjunction with a combination of other "soft" field marks. One newly discovered structural feature is a difference in "loral angle", which is a measure of how high the eye is positioned above the extension of the gape. Long-billeds have a more acute loral angle than Short-billed, giving the former a straighter-looking supercilium and the latter a more arched supercilium. These structural differences are speculated here to be an evolutionary manifestation of their fundamentally different habitat preferences on breeding grounds.
One notorious challenge in avian field identification is the dowitcher complex, which is composed of the Long-billed Dowitcher and three subspecies of the Short-billed Dowitcher. These species are superficially so similar that, up until the 1950s (Pitelka, 1950), the taxonomy of dowitchers was not yet agreed upon due to the difficulty in field identification (see Jaramillo et al. 1991 for a synopsis of the historical changes in taxonomy). Part of this confusion stems from seasonal variability in dowitcher plumages and from subspecific variation in Short-billed. However, it is now agreed that dowitchers can be divided into two species, the Long-billed (Limnodromus scolapaceus) and Short-billed Dowitchers (Limnodromus griseus), the latter of which consists of three subspecies: L. g. griseus, which breeds in the Maritime provinces, L. g. hendersoni which breeds in Alberta and Manitoba, and L. g. caurinus, which breeds in southwestern Alaska (Pitelka, 1950; AOU, 1957).
Significant advances in the field identification of dowitchers have been made over the last few decades by numerous field ornithologists, especially from the birding community (Wilds and Newlon, 1983; Hayman et al., 1986; National Geographic Society, 1999; Kaufman, 1990; Jaramillo et al., 1991; Paulson, 1993; Chandler, 1998; Sibley, 2001). For example, it is now common knowledge that the flight and alarm calls of the two species are different enough that voice can be used as a very reliable and easy field mark. There have also been considerable advances in the field identification of dowitchers in fresh alternate and juvenal plumage (Paulson, 1993; Jaramillo et al., 1991). However, silent birds in basic and worn alternate plumage still represent identification challenges. In such plumages, side-by-side comparisons are still identification challenges for experienced birders. To make matters worse, many birders, including seasoned birders, rely on habitat preference as a "field mark". Not only is this an useless field mark during migration when both species occur side-by-side and in habitats that they are not "expected" to be in, but it also masks any information regarding local movements of dowitchers between habitats. It is our firm belief that the accuracy of many dowitcher surveys during late summer through winter might be compromised by misidentifications.
Here, we present a set of new field marks that greatly aid in the identification of silent dowitchers. In particular, based on nearly a thousand hours of combined field observations and photographs of live birds, we show that the structural placement of the eye in relation to the bill systematically differs between the two species. The eyes on Short-billed Dowitchers are placed at a higher angle above the bill than on Long-billed Dowitchers. This difference in angle gives the two species distinctly different facial gestalts and also greatly aids in the identification of dowitchers in photographs, which are of course silent. We will also refine various plumage field marks for separating the two species. Subspecific variation will also be discussed. Finally, we speculate on the implications of our newly discovered field mark in the context of dowitcher taxonomy.
The results presented below are based on almost a decade of observing dowitchers at all seasons (in California, Massachusetts, and Texas). We also examined 100 + photographs and numerous specimens in the Museum of Vertebrate Zoology at UC Berkeley, the Los Angeles County Natural History Museum, and the Harvard Museum of Comparative Zoology.
2. Advances in structural field marks
A number of structural differences have been proposed to exist between the two dowitcher species. For example, as implied by the name, Long-billeds generally have longer bills than Short-billeds (Fig. 1). However, there appears to be considerable within-sex and inter-sex variation in bill length. Long-billeds having bill lengths at the long end of the spectrum (presumably females) are generally reasonably easy to separate from Short-billeds having bill lengths at the short end of the spectrum (presumably males), but unfortunately, the situation is usually not so ideal because, in many cases, the majority of dowitchers have overlapping bill lengths.
Other structural differences that have been mentioned in the literature include wing length and tarsi length (Pitelka, 1950). Pitelka showed that Long-billeds have slightly shorter wings but longer legs than all subspecies of Short-billed. Jaramillo et al. (1991) pointed out that Short-billeds may have a slightly longer primary projection, that is, the distance to which the primary feathers project beyond last tertial. In general, we have found these differences in primary projection (Fig. 2 and Fig 3) and tarsi length (Fig. 1) to be fairly robust. However, extreme care must be taken in using these field marks. Wear in the feathers or molting effects limit the use of primary projection as a fail-safe field mark (Fig. 4). In addition, wing length (and presumably primary projection length) exhibits some subspecific variation in Short-billed (Paulson, 1995). Differences in tarsi length are also subtle and may be of limited use in lone individuals that happen to be wading in water. However, the longer tarsi of Long-billed are noticeable in side-by-side comparisons (Fig. 1).
To build on these structural field marks, we embarked on a field study of dowitchers that collectively amounted to over 800 hours. We began to have a hunch in year 2001 that there might be subtle but distinct differences in bill and facial structures between the two species. In brief, the supercilium on Short-billeds tends to have an arched shape in all plumages, but that in Long-billeds tends to be straight. We also noticed that the bill on Short-billeds droops at the outer third of the bill, whereas that on Long-billed tends to be straighter. These features appear to give Short-billeds and Long-billeds very different facial expressions. Our attempts to test and quantify these qualitative hunches are outlined below.
2.1 Loral angle, supercilium and structural placement of the eye
We studied numerous dowitcher photographs, which included our own photographs as well as others from the literature and the internet. In an attempt to quantify "gestalt" or "jizz" features, such as facial expressions, we defined a parameter which we term here the loral angle (Fig. 5). Given a profile shot, this is the angle between an imaginary extension of the gape of the bill towards the back of the head and the line connecting the gape of the bill with the center of the birds iris (Fig. 5). An important feature of the loral angle is that it is a scale-independent variable, which makes it especially useful when studying birds in the field or in photographs (absolute bill length, on the other hand, requires that all photographed birds be scaled to the same size).
Loral angles were measured only on birds whose identity was unequivocally known based on a combination of plumage field marks (and sound, when available), which we later in this manuscript. Individuals of all plumages and ages were sampled in this study. Loral angles were measured only on individuals photographed as side-on profiles. We tried to avoid oblique views, such as head-on or top down views, because these views, when projected onto a two-dimensional plane, respectively, over-estimate and under-estimate loral angles. Despite our attempts to avoid oblique views, artifacts of projection probably could not be completely eliminated hence the variance in our measurements is a maximum estimate of the true variation of loral angle within a given species.
Figure 5c shows the results of our study. It can be seen from Figure 5c and various photographs (Fig. 1, Fig 2, Fig 3, Fig 4, Figs 5a, 5b) that Long-billeds have loral angles systematically smaller (more acute) than that of Short-billeds. There is a small overlap between the two species, but as stated in the previous paragraph, part of this overlap might be due to an artifact of projection. This difference in loral angle appears to be a manifestation of the fact that the eye on Short-billed is placed just slightly higher on the head than the eye on Long-billed. We are well aware that measurements of loral angles in the field on moving birds may be almost impossible. What is important, however, is how this subtle structural difference influences facial expression. The higher placement of the eye on Short-billed forces a more arched supercilium while that on Long-billed is straighter (Fig. 1, Fig 2, Fig 3, Fig 4, Figs 5a, 5b). The supercilium on Short-billed also tends to be slightly wider in the front, giving it a somewhat flared appearance, which accentuates the arched shape. In Long-billed, the supercilium is thinner, which further accentuates the straightness of the supercilium. These features are also apparent in head-on views. On Long-billed the superciliums meet at the forehead with a sharp straight-edged V (Fig. 6a). On Short-billed, the inside of the "V" (e.g., the dark median crown) comes to a more tapered point at the forehead (Fig. 6b).
As far as we know, the difference in loral angle has never before been mentioned in the literature. This structural field mark holds in all plumages and ages. With experience, it can even be discerned on birds in flight. We have independently field-tested this method over the past two years and have found this field mark to be remarkably robust, particularly in picking out a lone Short-billed within a flock of Long-billeds or vice versa. We caution, however, that one should never base identification on one field mark alone. Facial expression and loral angle at any given snapshot in time can vary depending on the birds posture or behavior (for example, facial expression can change when individuals open their mouths, feed or preen). It is the overall or average impression that should be considered in conjunction with additional field marks, which we will discuss below.
2.2 Bill shape
During our extensive studies, we also discovered that the two species have subtly different bill shapes (Fig. 1, Fig 2, Fig 3, Fig 4, Figs 5a, 5b, Figs 6). On Short-billed, the bill begins to gently curve downward about one third of a bill length away from the tip. This gives Short-billed a somewhat down-curved bill shape. On Long-billed, the bill is typically very straight along most of its length. The straight supercilium, more acute loral angle, and typically longer bill of Long-billed collectively serve to accentuate the overall very straight appearance of the bill. Again, we caution that bill shape can vary even on the same individual. Both dowitcher species have tactile mandibles and thus can temporarily alter the shape of their bills when feeding or probing.
2.3 Body shape
The two dowitchers differ in terms of their body shapes (see for example side-by-side comparison in Fig. 1 and schematized silhouettes in Fig. 7). The lower back on Long-billed has a slight indentation or concavity. This is probably a manifestation of the tertials being pushed up slightly by the tail in Long-billed. As a consequence, Long-billed tends to have a longer tail, a more tapered or attenuated rear, and a slightly "cocked" tail appearance (B. Small, pers. comm.). In Short-billed, the lower back is less indented and appears to have a stouter rear and less of a "cocked" tail appearance. This difference in body shape of course should not be used alone for identification. However, it can be used as an aid to quickly pick out a lone Long-billed in a Short-billed flock, or vice versa.
2.4 Head shape
Head shape is another structural field mark that can be very helpful when scanning dowitcher flocks. Long-billed has a slightly shallower forehead while Short-billed has a steeper forehead (Fig. 1, Fig 7). It is possible that these differences may be related to the different placements of the eye and structure of the supercilium among the two species. Indeed, the steep forehead, arched supercilium, and higher structural placement of the eye in Short-billed, gives it a generally distinctive facial expression. The shallow forehead on Long-billed helps to accentuate its straighter supercilium.
3. Ageing, molt, migration, and biology
3.1 Overall molt strategy in the context of migration
The first step in using plumage field marks is to determine age and molt stage. Individuals in their first several months of life are characterized by juvenal plumage. Juvenal plumage is characterized by pale to slightly buffy underparts, pale-edged feathers on the upperparts, and uniquely marked scapulars, coverts and/or tertials. Over most of the lower forty-eight states, juveniles begin arriving in August although juvenile Short-billeds arrive earlier. In Texas, juvenile Short-billeds arrive by 20 August whereas juvenile Long-billeds begin arriving at the very end of August and early September. The first juveniles to arrive are characterized by very fresh plumages (Fig 8, Fig 9). The juvenal plumage in both species will persist into late September and early October. By late October and early November, juveniles have largely molted into basic plumage, although remnants of juvenal plumaged tertials may still remain.
Fresh alternate plumage in both species is characterized by dark mantle feathers edged by bright buff fringes. In adults, molt into alternate plumage begins in early March on wintering grounds and involves the gradual replacement of tertial, covert and scapular feathers over the course of the next two months (molt in first year birds discussed in the next section). Molt into fresh alternate plumage is usually complete by late April such that northward bound migrants of both species are mostly in full alternate plumage. However, the timing of molt is not perfectly synchronous. For example, in April, any given flock of northward bound dowitchers of both species may contain individuals at various stages of molt into alternate plumage (Fig. 10). By May, however, both species are in full alternate plumage (Fig. 11).
Basic plumage in both species is characterized by gray upperparts and whitish underparts (Fig. 12 and Fig 13). Molt from alternate to basic (in adults) begins in both species as early as late July, but most of this early molt involves slight changes to head or underpart coloration and the replacement of a few covert feathers. First fall arrivals of adults in both species are largely in full, but worn alternate plumage (Fig. 1 and Fig 2). Most of the transition from alternate to basic plumage is completed on wintering grounds and is largely complete for most birds by the end of August or early September. One difference, as pointed out by Putnam (2005), is that Long-billeds molt their primaries during transit whereas Short-billeds wait until they arrive at their wintering grounds to molt their primaries. Examples of wing molt are shown in Fig14 and Fig 15.
Putnam argued that for birds that can be unequivocally demonstrated to be migrants, the differing molt strategy can be a reliable field mark (this field mark is unlikely to be useful anywhere along the Pacific, Gulf or Atlantic coasts, which are not only migration stopovers but also wintering grounds for Short-billeds). We caution, however, that more tests of this field mark are probably needed. For example, in mid-July, along the upper Texas coast, some of the first Short-billeds to arrive (or pass through) have already lost their inner primaries (Fig. 14), suggesting that molt strategy may be more complicated than suggested by Putnam. It is possible that the apparent difference in molt strategy suggested by Putnam is due to the fact that Short-billed migrates south through the lower forty-eight states slightly earlier than Long-billed. In the Pacific Northwest and along the central California coast, Short-billeds (caurinus) arrive respectively in mid- and late June, while Long-billeds arrive respectively in early to mid-July (Paulson, 1993; Roberson, 2002). In Illinois, Short-billeds begin arriving in early to mid-July and Long-billeds much later (Robinson, 1996). Along the upper Texas coast, Short-billeds (hendersoni) arrive in early July while Long-billeds arrive in mid- to late July (Lee, pers. obs.; Eubanks et al., 2006). In New England, Short-billeds (presumably mostly griseus) begin passing through in early July while Long-billeds begin passing through in mid- to late July (Veit and Petersen, 1993). We conclude that rough differences in molt strategy (Putnam, 2005) may be a helpful field mark, but should be used only in conjunction with other field marks. An additional complication in the molt "field mark" is the molting strategy of first year birds, which we discuss below.
3.2 Peculiarities of first year birds
Our above discussion of the pre-basic molt pertained only to adult birds, at least in the case of Short-billed Dowitchers. Paulson (1993) suggested that Short-billeds may take 1 year to mature into adults and that, during their first spring, they may actually molt directly into 2nd basic plumage. Based on experience from the Pacific Northwest, Paulson also suggested that a few of these first year Short-billeds may summer locally on their wintering grounds. Our surveys along the upper Texas coast (2002-2005) confirm Paulsons speculations. Large numbers (possibly up to 400) of Short-billeds (hendersoni) remain through the summer along the upper Texas Coast (Fig. 6b, Fig 16, Fig 17, Fig 18; Lee pers. obs.; Weeks, pers. obs.). For example, in June along the upper Texas coast, 95% of these summering individuals are in fresh basic plumage although many may show one alternate tertial or a few alternate covert feathers (Fig. 16, Fig 17). The remaining ~5% are birds that have only partly molted into alternate plumage; such birds show a tint of red or orange on the undersides, but have a number of basic-like covert feathers (Fig. 16, Fig 17). An important feature of all of these summering birds is that, although they are mostly in fresh basic plumage, their outer primaries appear to be pale brown, and hence worn. In early July, there is a rapid increase in the number of Short-billeds along the upper Texas coast, but curiously, nearly all of these are in worn alternate plumage. The primaries on these fall arrivals tend to be darker, and hence fresher, than those on individuals that have summered locally.
These observations seem to corroborate the suggestion by Paulson that summering individuals are first year birds. Additional circumstantial evidence in support of this suggestion comes from our surveys of shorebirds at the Houston Audubon Societys Bolivar Flats, an extensive mudflat region that serves as an important migratory stopover and wintering ground for many shorebirds along the Texas coast. During June 2004, approximately 150 Short-billed Dowitchers summered at Bolivar Flats. In contrast, only ~30 were found at this same locality in June 2005, which followed an unusually poor shorebird nesting season the previous year as evidenced by the fact that during the fall of 2004, very few juvenile Short-billeds (~<2 % of the population) were observed at Bolivar Flats. Many more juvenile Short-billeds were observed in the fall of 2005.
It seems likely that Short-billeds take one year to mature fully in terms of plumage. During their first spring, they molt directly into basic plumage. Many of these individuals over-summer on their wintering grounds along the Gulf Coast (hendersoni). A few may summer locally in the Pacific Northwest, central California coast, and the Atlantic coast (Zeranski and Baptist, 1990; Paulson, 1993; Veit and Petersen, 1993; Sibley, 1997; Roberson, 2002), but total numbers are much lower than on the Gulf Coast. Interior summer records away from breeding grounds are rare, but not non-existent. For example, Patten et al. (2003) report several June records of Short-billeds at the Salton Sea in southern California and Robinson (1996) reports one June record for southern Illinois; these birds could potentially represent summering individuals.
The implications of delayed maturation of Short-billeds are that molt timing and systematics may be much more complicated than Putnam (2005) suggested (see above). First year Short-billeds in July show highly worn outer primaries but actually have fresh, full grown, inner primaries indicating that they have already molted and replaced their flight feathers by mid-summer. This is illustrated in Figure 18 by the darker coloration of the full-grown inner primaries, which contrasts with the paler brown outer primaries on flying or preening birds. By contrast, alternate-plumaged fall arrivals of Short-billeds on the upper Texas coast are only then beginning to molt their inner primaries as evidenced by distinct "windows" in their secondaries and primaries (Fig. 14). This means that these first year birds begin molting inner primaries much earlier than the adults. If some of these Short-billeds summered in interior localities or to the north, then it is likely that some southbound Short-billeds will already have started molting their primaries. Thus, care must be taken to age the birds before molt timing is used as a field mark.
The question that follows is whether Long-billeds behave similarly in terms of plumage and maturation. In stark contrast to Short-billeds, summer records of Long-billeds away from breeding grounds seem to be rarer in general. Small numbers regularly summer along the coast in the Pacific Northwest, but total numbers are nowhere near those of summering Short-bills (Paulson, 1993). Similarly, we have recorded a few Long-billeds in basic-like plumage along the upper Texas coast in late May and early June although none of these stayed through June. One place where summering Long-billeds may be more common is in the Salton Sea. Patten et al. (1993) report that non-breeding Long-billeds regularly summer at this spot each year, the highest summer count being 40 on 20 June 1991. Clearly, much more work is needed to clarify the migratory and molt behavior of first year Long-billeds. However, the above observations suggest that at least some Long-billeds may over-summer locally on wintering grounds. Are such birds first year birds, and if so, how does this complicate our understanding of molt timing