HARVARD UNIVERSITY «« LIBRARY OF THE Museum of Comparative Zoology / r •/ l^ i^Ai-a CG.v.p. icv^OL. LIBRARY ^^' 18 1968 HARVARD RELATIONSHIPS ^jniversity, IN THE CHARADRII (SHOREBIRDS): A TAXONOMIC STUDY BASED ON COLOR PATTERNS OF THE DOWNY YOUNG JOSEPH R. JEHL, JR. SAN DIEGO SOCIETY OF NATURAL HISTORY MEMOIR 3 1968 f Ted hWe. I "HtS _> -nif - MLIo. COuiP. ;<.wOL. LIBRARY OCT l8 1968 HARVARD R E L AT I O N S H I PS university IN THE CHARADRII (SHOREBIRDS): A TAXONOMIC STUDY BASED ON COLOR PATTERNS OF THE DOWNY YOUNG JOSEPH R. JEHL, JR. SAN DIEGO NATURAL HISTORY MUSEUM SAN DIEGO SOCIETY OF NATURAL HISTORY MEMOIR 3 1968 f SAN DIEGO SOCIETY OF NATURAL HISTORY MEMOIRS Memoir 3, pages 1-54, figures 1-31, frontispiece Issued September SO, 1968 Price: $3.50 a copy Frontispiece. Top to bottom. Day-old chicks of Micrapalama himantapus, Limnodromus griseiis, Arennrin iyiterpres, Limosn haemastica. Limosafedou. Watercolor by Ted Miley. CONTENTS INTRODUCTION 5 Acknowledgments 6 COLOR PATTERN AS A TAXONOMIC CHARACTER 7 Conservatism of plumage patterns 7 Experimental data 8 COLOR PATTERNS OF SHOREBIRD CHICKS 9 FAMILY JACANIDAE 11 FAMILY ROSTRATULIDAE 12 FAMILY HAEM ATOPODIDAE 12 FAMILY CHARADRIIDAE 12 Subfamily Vanellinae 12 Subfamily Charadriinae 13 FAMILY SCOLOPACIDAE 16 Subfamily Tringinae - 16 Subfamily Arenariinae 20 Subfamily Scolopacinae 20 Subfamily Calidridinae 23 FAMILY RECURVIROSTRIDAE 23 Subfamily Ibidorhynchinae 23 Subfamily Recurvirostrinae 23 FAMILY PHALAROPODIDAE 25 FAMILY DROMADIDAE 26 FAMILY BURHINIDAE 26 FAMILY GLAREOLIDAE 26 Subfamily Cursoriinae 26 Subfamily Glareolinae ■ 28 FAMILY THINOCORIDAE 28 FAMILY CHIONIDIDAE 29 RELATIONSHIPS IN THE CHARADRII 29 JACANIDAE, ROSTRATULIDAE 29 DROMADIDAE 31 HAEMATOPODIDAE, RECURVIROSTRIDAE, BURHINIDAE 31 Intrafamilial relationships 31 Interfamilial relationships 31 CHARADRIIDAE 32 Vanellinae 32 Charadriinae 32 Interfamilial relationships 34 GLAREOLIDAE 34 Intrafamilial relationships 34 Interfamilial relationships • 34 SCOLOPACIDAE .37 Tringinae - 37 Arenariinae 37 Scolopacinae 38 Calidridinae 39 Phalaropodinae 40 Interfamilial relationships 40 THINOCORIDAE, CHIONIDIDAE 40 PHYLOGENY 41 CLASSIFICATION 43 SUMMARY ......'.'.'..'.... .....U APPENDIX - References to descriptions and illustrations of downv young shorebirds 45 LITERATURE CITED ' ' 51 INTRODUCTION The avian order Charadriiformes is composed of three suborders: Charadrii, Lari, and Alcae. By far the largest and most diversified of these is the Charadrii (shorebirds or waders), in which seven superfamilies, 12 families, and over 200 species are currently recognized (Peters, 1934). Since the late Cretaceous this group has undergone an extensive degree of adaptive radiation, particularly in bill morphology, which equals or surpasses that found in any other avian suborder. The major shorebird lines, currently recognized as families, are dis- tinguished largely by differences in bill shape. In most cases the allocation of species to families, subfamilies, and genera is well established. How- ever, the interrelationships of these higher taxa are inadequately known, as shown by the variety of classifications that have been proposed in recent years (e.g., Lowe, 1931b; Peters, 1934; von Boetti- cher, 1934; Mayr and Amadon, 1951; Verheyen, 1958a; Yudin. 1965; Brodkorb, 1967). Most of these classifications have been based largely on P. R. Lowe's extensive studies of the Charadrii and their close relatives in the Gruiformes. Lowe studied many characters, but he placed greatest emphasis on the morphology of the bill and skull. Con- sequently, his work was primarily valuable in emphasizing lines of divergence, but not in clari- fying interrelationships within the Charadrii. In this study I have attempted to reassess inter- relationships in the Charadrii by analyzing vari- ation in the downy plumage. Plumage patterns of adult birds probably have been used more often than any other character in inferring relationships at the generic level, and sometimes above. How- ever, chick patterns have been largely ignored, even though in some taxa (e.g., Anatidae: Delacour and Mayr, 1945; Tetraoninae: Short, 1967; Podicipe- didae: Storer, 1967) they have proved valuable in suggesting or confirming relationships. Recogni- tion of the potential taxonomic value of shorebird chick plumages is not original. It was first noted by Lowe (1915) and subsequently by Ticehurst (1933) and Verheyen (1958a). Unfortunately, these work- ers, perhaps handicapped by inadequate material, made no more than preliminary attempts to study chick plumages, and the results of their in- vestigations played only a minor and sometimes inconsistent role in their classifications. Since downy plumages have not been fully utilized pre- viously, and since, in general, there is no important correlation between adult and chick plumage pat- terns, except in the plover genus Charadrius. (see p. 32), the downy plumages can be used as an essen- tially independent means of evaluating relation- ships in this suborder. A method for investigating the taxonomic significance of a single character or character complex has been outlined by Bock (1960: 363-364). Briefly: 1) the occurrence, structure, and variation of the character is studied, 2) its functional significance is determined, and 3) its evolution is postulated. As Bock has emphasized (p. 364), un- derstanding the functional significance is the most important part of the investigation, for with this knowledge "one can estimate the selection forces which were operative during the evolution of the character. A knowledge of the selection forces is essential because while it is possible to outline the phylogeny of a structure without knowing the selection forces, it is impossible to understand its evolution without them." With this information the taxonomic value of the character can be eval- uated. The downy plumage of shorebird chicks is best considered as a character complex consisting of four partially independent characters: color, color pattern (or pattern), feather structure, and pat- terning of individual feathers. For reasons dis- cussed below I have placed greatest emphasis on the relationships indicated by color pattern. The color of young shorebirds seems to be strongly correlated with that of the substrate in nesting areas (Cott, 1957: 124-131); for this reason it is usually of little taxonomic value. Feather structure and patterning of individual feathers, in general, show too little variation to be of wide usefulness; however, some significant exceptions occur, and those considered ta.xonomically important are dis- cussed. Some types of feather structure may result from convergence; for example, the down of arctic species is long and densely barbed, whereas that of temperate-zone forms is shorter and more sparsely barbed. Convergence in patterning of individual feathers, e.g., the white-tipped dorsal down of Thinocarus, Oreopholus, Calidris, and Cursorius, also occurs. JOSEPH R. JEHL, JR. In this study I examined chicks of 109 of the 201 shorebird species and photographs or descriptions of 53 additional species. The number of genera examined is a more important criterion of com- pleteness, however, because there is little in- trageneric pattern variation. Chicks of 41 of the 56 genera recognized here were available; descriptions or photographs were available for all others except Micropurra (Jacanidae), Plurl(niellu>< (Charad- riidae), and Prosobonia (Scolopacidae). The ab- sence of significant intraspecific plumage variation obviated the use of large series, but several speci- mens of each species were examined if possible, for even in the most carefully prepared skins the patterns are easily disarranged. For maximum value in future studies field workers are encour- aged to prepare chicks of all ptilopaedic species as shown for Chamdriua semipalmatus (Fig. 7). Specimens were borrowed from the British Mu- seum (Natural History), Carnegie Museum, Chi- cago Natural History Museum, Museum of Com- parative Zoology, Museum of Vertebrate Zoology, National Museum of Canada, Royal Ontario Mu- seum, George M. Sutton, and David F. Parmelee. In addition I examined the entire collections of shore- bird chicks in the following institutions: Ameri- can Museum of Natural History, The University of Michigan Museum of Zoology, the San Diego Natu- ral History Museum, and the United States Na- tional Museum. Living chicks of 19 species were also studied. Misidentified specimens pose a potentially serious problem in a study of this type. Several were detected once the degree of intraspecific and intrageneric variation was known; in each case the confusion was with a sympatric species. Possible errors in identification to species, though not to genus, may still exist for species in which only a single chick was examined. However, from consid- eration of previously published descriptions, geo- graphic locality of the specimen, reliability of the collector, and the fact that chicks of most species are distinguishable, the possibility seems unlikely. Observations on the functional significance of chick plumages were made at Churchill, Manitoba, Canada, in the summers of 1964, 1965, and 1966. ACKNOWLEDGMENTS I am indebted to Harrison B. Tordoff, Richard D. Alexander, Claude W. Hibbard, and Robert W. Storer for their assistance and encouragement during this study. The manuscript has benefitted from their criticisms and from those of Kenneth C. Parkes. I am grateful to the following curators for allow- ing me to examine, often for extended periods, specimens in their care: Dean Amadon, Emmett R. Blake, Jon C. Barlow, W. Earl Godfrey. Ned K. Johnson, J. D. Macdonald, Kenneth C. Parkes, Raymond A. Paynter, Robert W. Storer, George M. Sutton, George E. Watson, and Richard L. Zusi. Keith A. Hindwood, Gordon L. Maclean, Peter Roberts, Robert W. Storer, and G. M. Storr kindly provided unpublished photographs or descriptions of species that I otherwise would have been unable to consider. Many persons aided my field work in Churchill, Manitoba, but special thanks are due R. B. Angus, Duncan R. Mackenzie, David F. Parmelee, Mrs. I. H. Smith, and my field companions, Charles G. Yarbrough and D. J. T. Hussell. The following persons extended various courte- sies: Alfred M. Bailey, Pierce Brodkorb, John Bull, Charles T. Collins, Paul DeBenedictis, Eugene Eisenmann, J. Alan Feduccia, Norman L. Ford, Frank B. Gill, John P. Hubbard, C. Jouanin, A. S. Loukashkin, Sheldon Miller, Bertram G. Murray, Jr., William C. Preston, Charles G. Sibley, and Alexander Wetmore. Special thanks are due John Tottenham for his careful execution of the pattern diagrams and Ted Miley for the colored frontis- piece. This research was supported, in part, by the Frank M. Chapman Memorial Fund of the Ameri- can Museum of Natural History, The University of Michigan Horace H. Rackham Graduate Student Research Fund, a National Science Foundation Summer Fellowship for Teaching Fellows, and a National Science Foundation Assistantship in Sys- tematic and Evolutionary Biology administered by The University of Michigan Museum of Zoology (Grant GB-3366). RELATIONSHIPS IN THE CHARADRII COLOR PATTERN AS A TAXONOMIC CHARACTER It is not at all obvious why chick patterns should be expected to yield satisfactory taxonomic infor- mation. Delacour and Mayr<1945: 4) justified their use by terming them "non-adaptive" characters. However, in nearly all animals a high percentage of the annual mortality occurs among the very young, and any characteristics that minimize this loss will be of selective advantage. In shorebirds, and per- haps in all groups with precocial young, the plu- mage patterns are cryptic or disruptive and func- tion to reduce the chicks' conspicuousness (Cott, 1957:54,60-61, Fig. 14, PI. 19). Thus, pattern p^rse must be considered adaptive even though several patterns may appear to be of similar selective value in a particular habitat (see below ). Several persons have suggested to me that chick patterns might also aid the parents to recognize their young, but this does not seem to be the case in any species I have studied in the field (Charadrius semipalmatus, Pluvialis dominica, Numenius phaeopus, Limosa haemastica, Tringa flavipes, Limnodromus griseus, Gallhiago gallinago, Cal- lidris pusilla. C. minutilla. C. alpinn, Micropalama himantopus, Phalaropus lobatus). Chicks of these species are often obscured from the parents' sight by vegetation, and maintenance of the family group appears to be accomplished primarily, if not entirely, through vocalizations. The absence of "character displacement" in chick plumages of closely related sympatric species, and the observa- tion that some shorebird species may attend other than their own young (Parmelee, Greiner, and Graul, 1968: 20-21; Jehl, pers. obs.l, further in- dicates that visual recognition may be of minor importance. CONSERVATISM OF PLUMAGE PATTERNS In spite of their adaptive role, downy patterns appear to be highly conservative and to reflect phylogenetic relationships and not environmental difl'erences for the following reasons: 1) The patterns are restricted to the Charadrii and do not occur in groups thought to be closely allied to this suborder. 2) Patterns show little variation within species, genera, and sometimes within higher taxa. 3) Patterns are similar in closely related genera. 4) The young of closely related species occupying different habitats possess similar downy patterns. For example, in North America, Haematopus pal- liutus nests on sandy beaches, whereas H. bach- mani nests on rocky shores; the patterns, though not the coloration, of their chicks are identical. In Europe, H. ostralegus nests "on or near shore, on shingle beds, among rocks, in sand dunes, or among grassy banks" ( Witherby et al., 1940: 416); there is no intraspecific pattern variation in this species. 5) The young of distantly related species oc- cupying the same habitats at the same time possess differing downy patterns. For example, in wet marshes at Churchill, Manitoba, in early July, the similarly-colored grayish chicks of Numenius phaeopus, Limosa haemastica, and Tringa flavipes are present; three distinct color patterns are repre- sented by their plumages. In the same marsh occur the brownish chicks oi Limnodromus griseus, Gal- linago gallinago, Calidris minutilla, C. alpina, C. pusilla, Micropalama himantopus, and Phalaropus lobatus; three additional color patterns are repre- sented. In addition, young of Branta canadensis. Anas actua, Coturnicops noveboracensis. and Lag- opus lagopus (four additional patterns) are often present. Many additional examples could be ad- vanced. I know of no instance in which the presence of similar or identical color patterns should be attri- buted to convergence. In many groups patterns are so complex that the independent evolution of sim- ilar patterns seems highly improbable. Further- more, the precise configuration of patterns is not usually obvious in living chicks; for example, "dia- mond-shaped" patterns on the lower back of cur- lews and godwits can rarely be discerned, but appear as a random mixing of dark and light down. Thus, it cannot be argued that similarities reflect selection for specific patterns. An exception must be made in the case of pattern loss, which has occurred in many unrelated forms, notably species nesting along mountain streams or on uniformly colored substrates, e.g., Ibidorhyncha, Phegomis, Heteroscelus, Cladorhynchus. I would conclude from the above that shorebird chick patterns are conservative, that different pat- terns may be adaptive in similar habitats, and that JOSEPH R. JEHL, JR. some patterns may be adaptive in a variety of habitats. EXPERIMENTAL DATA It would be desirable to provide experimental evidence for the adaptive equivalence of downy plumage patterns in different habitats. Unfortu- nately, selection experiments with nidifugous chicks are extremely difficult to perform under natural conditions; the survivorship of individual species cannot be measured precisely, nor can sufficient natural predation on the chicks be ob- served. However, a crude predation experiment was attempted. Specimens of seven species of waders (Phivialis dominica, Tringaflavipes, Actitis macularia, Scotopaj: minor, Gullinugo galUnago, Calidris alpina. Phalaropus lobatus), and one grouse (Lagopus lagopus). each with a different color pattern, were photographed on the substrates in several different types of habitat (Fig. 1). At Churchill, Manitoba, these included a wet marsh, a dry marsh, and a lichen-covered hillock; at Ann Arbor, Michigan, a deciduous woods. Photographs were taken in color, from a uniform height, and within 30 feet of a nest of one of the species or in areas where chicks of at least one species were known to occur regularly. From one to eight speci- mens were included in each picture. Twenty-one such photographs were then viewed by ten persons, in most cases graduate students in ornithology, previously familiar with downy plumages of the species involved. These "predators" examined each slide for a period of 20 seconds, during which they were asked to indicate the approximate location of chicks in the order in which they were found by marking "1", "2", etc., on prepared forms. The short viewing time insured that the less-conspicuous chicks would be overlooked. The results of this experiment could not be quantified and therefore must be regarded as in- conclusive. Variation in conspicuousness of chicks resulting from unevenness of topography, relative size of the specimens, "make" of the skin, and position of the specimen in the photograph, and so forth, precluded any constancy between pictures. However, by determining for each picture which chicks were not found or were found only with great difficulty, I subjectively evaluated the con- Fig. 1. Specimens of downy young photographed in a dry marsh at Churchill, Manitoba. Left to right: Phaliimpus lobatm. Tringa flavipes, Scolopcue minor. RELATIONSHIPS IN THE CHARADRII spicuousness of chicks in each habitat. It appeared that chicks of Tringa flavipes, Actitis macularia, Gallinago galUnago, Calidris alpiua, and Phala- ropus lobutus were relatively inconspicuous on all substrates. The Pluvialis dominica chick was highly conspicuous in the dry tundra marsh and in the deciduous woods, the Scolopa.v minor chick on the lichen-covered hillock. The Lagopus lagopus specimen was inconspicuous on all substrates at Churchill, but was not tested in deciduous woods. Thus, the results of this artificial experiment sug- gest that some patterns may be concealing in a variety of habitats and that different patterns may be concealing in the same habitat. Workers who may wish to conduct similar experiments should consider the following improvements in technique: 1) specimens should be approximately the same size and should be prepared to simulate the normal crouching position of chicks; 2) legs of specimens are particularly conspicuous and should be re- moved or folded to lie under the body; 3) photo- graphs should be taken on black-and-white and on color film so that the concealing effects of pattern as well as coloration can be differentiated; 4) the effect that the specimen's position in the photograph has on its conspicuousness should be examined. COLOR PATTERNS OF SHOREBIRD CHICKS Chick color patterns are described below for each genus and the extent of intrageneric variation is noted; representatives of many genera are illus- trated and the major variations in pattern are diagrammed. In most cases the diagrams are com- posites made from the examination of several specimens, but whenever possible they were drawn from living chicks. In interpreting the diagrams, greatest emphasis should be placed on the promi- nent back and head markings; minor markings on the face, back, and wings are variable and seem to be of no taxonomic importance. Before making interpretations the reader should compare the diagrams with the photographs, for complete re- liance on the diagrammed patterns could lead to error; chicks of Lintnua and Calidris, for example, are far more different than the diagrams alone can indicate. No attempt has been made to provide detailed descriptions of the downy plumages for individual species, since adequate descriptions of most species already exist. However, original descriptions are given in the Appendix for species that have not been described previously or for which the pub- lished descriptions are inadequate. The classification followed is basically that of Peters (1934), the last revisor of the entire subor- der; minor departures are noted in the text. Since the appearance of Peters' classification several generic mergings have been proposed. I advocate those discussed below. 1) Bock's (1958) generic classification of the Charadriidae is followed, except, as discussed be- low, Oreopholus is considered distinct from Eu- dromias. 2) Aechmorhynchus is merged with Prosobonia for reasons that will be discussed elsewhere (Zusi and Jehl, in prep.). 3) Pseudototanus is merged with Tringa. This merging has been suggested by several workers (e.g., Vaurie, 1965); I consider the characters listed by Hume (1876: 345-347; 1878: 489) insufficient to support generic distinctness. 4) I follow De Schauensee (1966) in merging Chubbia with Gallinago. 5) Peters (1934) and others have classified the woodcocks in two genera, Scolopax and Philohela, the latter containing only one species, minor. Char- acters used to differentiate Philohela include: 1 ) the more rounded tail, which contains 14 instead of 12 rectrices and in which the outer rectrices are narrowed, 2) the attenuation of the outer three primaries, and 3) the eutaxic condition of the secondaries (Ridgway, 1919; Miller, 1915). In addi- tion, the courtship displays oi Philohela differ from those of Sco/opax nisticola. In my opinion these points do not justify the continued generic recognition of Philohela. Vari- ation in tail shape and in the number of retrices also occurs, and to a much greater extent, in Gallinago (Baker, 1929: 265, PI. 3), a genus that is usually assumed closely allied to Scolopax. The configuration of the primaries is associated with courtship behavior, the attenuated feathers being used to make mechanical sounds during the court- ship flight (Pettingill, 1936: 288). Philohela is the only reported eutaxic sandpiper, but variation within other families, genera, or even a single individual is known (Miller, 1915, 1924; Verheyen, 1958b). The taxonomic significance of this condition is debatable, but since eutaxy is usually associated with a short-winged condition in other groups, it is 10 JOSEPH R. JEHL, JR. Fig. 2. Pattern diagrams of: A. Actxyphilomis africana; B. Rostratula benghalensis; C. Haematopus palliatus; D. Vanellus duvaucelii: E. Charadrius semipatmatus; F. Pluvialis dominica: G. Numenius phaeop^is; H. Limosa haemastica: I. Tringa flavipes: J. Xenus cinereus: K. Actitis macularia; L. Limnodromus griseus: M. Gallinago gallinago; N. Scolopaa: minor: 0. Calidris ptilocnemis: P. Himantopus himantopus; Q. Phalaropus lohatus: R. Burhinus oedicnemiis. RELATIONSHIPS IN THE CHARADRII 11 not especially surprising to find it appearing in this relatively short-winged group of sandpipers. The significance of the courtship differences cannot be evaluated, for courtship displays have been de- scribed for only minor- and rusticola (Pettingill, 1936; Steinfatt, 1938). However, the differences between these species are no greater than differ- ences within Gdlliucgo or Calidris. All the differences noted may be correlated with breeding behavior. The epigamic role of the pri- maries has been noted; the rectrices are also used by the males in courtship (Pettingill, 1936: 295-296) and by the females in distraction displays (Pettin- gill, 1936: 332; pers. obs.); a possible functional relationship between eutaxy and the attenuated condition of the primaries deserves further study. Although diflFerences in breeding behavior indicate species rank, they do not necessarily warrant the recognition of higher taxonomic categories. I sug- gest that it is more important to emphasize the obvious close relationship of Philnhela and Scolo- pax by combining both into a single genus than to emphasize difTerences associated with reproduc- tive behavior by maintaining P/;//o/(e/« as a mono- typic genus. 6) Calidris is used to include Crocethia, Ereu- netes. and Eroliu (cf. British Ornithologists' Union, 1952; Kozlova, 1962; Holmes and Pitelka, 1964 1. 7) I accept the merging of Steganopus and Lo- bipes with Phularopus. long advocated by most European workers. 8) Orthorhamphus is considered congeneric with Esacus. There has been little agreement on the number of valid burhinid genera. Hartert (1924: 262) lumped Orthorhamphus with Esacus. Mei- nertzhagen (1924) went farther, lumping these genera and Oedicnemus with Burhinus. Baker (1929: 80) accepted the lumping of Oedicnemus and B^irhinus, but stated that "as Orthoramphus [sic] differs from Esacus quite as much as Burliinus does from that genus, it seems only consistent to have three genera, the alternative being to lump all the species in one genus." Peters (1934) followed Baker and retained three genera. Mayr (1938: 15), how- ever, not only found the differences between Or- thorhamphus and Esacus slight, but he also stated that "the two species are so similar that one is strongly tempted to unite them into one." Deignan (1945: 129) formally adopted this view, which Mayr (1949: 30), in turn, supported. This view may be correct, but to my knowledge no evidence support- ing conspecificity has been presented. Although the specific status of these forms is unresolved, it is evident that these long-billed burhinids form a morphological unit that is dis- tinct from the shorter-billed forms now grouped in Burhinus. I consider the Burhinidae to comprise two genera, Burhinus and Esacus, and Esacus to comprise two species recurvirosfris and magni- rostris. In the following discussion the number of speci- mens of each species examined is given in paren- theses; species for which only descriptions or pho- tographs were available are indicated by an aster- isk. Family JACANIDAE Microparra: capensis Actophilornis: ufricana (1), albinucha (1) Irediparra: galliyiacea* Hydrophasianus: chinirgus* Metopidius: indicus* Jacana: spinosa (1) Chicks of all species of jacanas are white ven- trally, yellowish-brown to chestnut dorsally. There is a distinct dorsal pattern (Fig. 2A). A brownish crown patch is bordered posteriorly and laterally by black postocular stripes that meet on the occiput and extend down the neck as a thin black line. In Irediparra (Fig. 26), Jacana. and both species of / '■■'..' -I /J Jslacoii,-/ Fig. 3. Actophilornis africana and A. albinucha. 12 JOSEPH R. JEHL, JR. Actophiloniis (Fig. 3) the stripe continues ante- riorly and forms a tiiin central stripe on the hind portion of the crown; in A. (ifricinui and Iredipurra, but not in A. ulbinuchd, there is an indistinct black supraorbital spot. A broad, brownish midback stripe is flanked by narrower whitish stripes which, in turn, are bordered by brownish stripes on the sides of the body; these lateral stripes are usually paler than the midback stripe. In A. africana, Iredipurru, and Jacinid the whitish stripes are outlined by a narrow black edging, but in A. albinucha, and possibly in Metopidius, the up- permost black edging is lacking. In the species examined the wings are brownish with a black stripe on the leading edge; there is also a thin dark line along the lateral surface of the thigh. Family ROSTRATULIDAE Rostra i u lu: benghalensis * Nycticryphes: semicollans* I have not been able to examine chicks of either species of painted snipe, but K. A. Hindwood has provided me with several excellent photographs (Fig. 26) from which I have diagrammed the dorsal pattern of Rostmtula (Fig. 2B). Hindwood (1960: 68-69) described i?os^rafwk chicks as follows: General colour of upperparts, including crown, light fawn; underparts slightly darker. A black stripe from the base of the upper mandible extends over the centre of the crown, where it opens out into an elongate-oval, to close again on the nape; the colour between the oval on the crown is light brown. A bright tan stripe, appro.ximately 1/8 inch wide, extends along the entire back, and is bordered by narrower black lines. A narrow black stripe commences at the base of the lower mandible and e.xtends through the eye along the entire length of the body about level with the small, undeveloped wings. A shorter black line, about half the length of the body, occurs on the sides of the abdomen below the wings. The young of Nycticryphes, judged by a very rough sketch (Pereyra, 1943: 180) are similar to those of Rostratula, but may have a solid crown cap. Family HAEMATOPODIDAE Haematopus: ostralegus (1), palliatus (10), bdch- m(nii{2), leucnpodits* ,fHli()i)insns* , (iter {!) Oystercatchers constitute a particularly well- defined wader group, but species within the group are not easily delimited. Four to nine species have been recognized (cf. Peters, 1934; Webster, 1943; Oliver, 1955). Here I follow the American Ornitho- logists' Union (1957) for the American forms, Pe- ters for the others. Webster (1943) has previously reviewed the Fig. 4. Haematopus palliatus and H. bachmani. downy plumages of oystercatchers. Nearly all of the forms that he studied were individually recog- nizable. Yet, a constant pattern, which he described as follows (pp. 40-41), occurs throughout the fam- ily. (11 Pure white down is present on more or less of the underparts in every form hut fuliginosns [bachmani and some chicks of H. p. frazari]. (2) Black down forms two parallel stripes on the back, a stripe at the top of the thigh, a dot on the lores, and a dot behind the eye; also spots and a median stripe on the pileum in some forms. (3) Some shade of drab down covers those areas of the upper parts which are not black, the throat, and (in the dark-bellied forms) most of the ventral surface. (4) Some of the drab down of the upperparts is tipped with a shade of buff. The pattern of//, palliatus diagrammed in Figure 2C may be compared with photographs of H. palliatus and //. bach rii atii (Fi^. 4). In both species there is a definite band of dark down on the occiput and nape. Family CHARADRIIDAE Subfamily VANELLINAE Vauellus: vayiellus (6), crassirostris. arniatus*, spinosus (1), duvaucelii (1), tectus (2), malabar- icus, albiceps, lugubris*, melanopterus* , coro- natus*. senegallus*, melanocephalus, super- cilinsus, gregnrius*, leucurus, cuyanus, chilensis RELATIONSHIPS IN THE CHARADRII 13 . V y r Fig. 5. Vanellus chilensis. V. duvaucelii, and V. tectiis. (2), resple}ide)is, ciuereuH*, iytdicus*, macrop- tenis. tricolor ( 1 ), in iles ( 1 1 There would appear to be relatively little vari- ation in color or pattern of lapwing young. Typi- cally, chicks are grayish-brown mottled with fine black markings dorsally, except for a white band on the nape (Fig. 2D|. In all species hul (/regariuH the mottling on the head is sharply demarcated post- eriorly by a solid black band from the posterior corner of the eye, or in some cases from the auricular region, to the occiput. A black band of variable prominence extends from the hip to the tail. Most species lack a definite back pattern, but black down forms an interrupted V in duvaucelii and tectiis, a solid midback stripe in npiyios^is, and a "broken black line down the center" in coronatus (van Someren 1956: 125). The venter is white in all species; vanellus. chilensifi, and iudicus possess a black pectoral band. Chicks of chilensia, tectum, and duvaucelii are shown in Figure 5. Subfamily CHARADRIINAE Charadrius: hiaficnhi (31), nemipalmatus (37), pla- cidus, dubius (5), icilsoniu (9), vociferus (33), melodus (10), thoracicus, pecuarius (3), sanc- taehelenae, tricollaris (1), alexandrinus (20), marginatus, peronii (2), venustiis*, collaris, bi- cinctus*, falklandicus (1), nwngolus*, les- chenaultii*, asiaticus*, veredus, rnodestus*, montamis (10), melanopsd), cinctus, nibricollis, uovaeseelandiae* This arrangement of plover species follows Bock (1958), except that nemipalmatus, sanctaehelenae, marginatus, and veredus are considered full spe- cies (see Vaurie, 1964). The basic downy pattern in Charadrius (Fig. 2E), like that of Vanellus, consists of a mottled crown patch separated from a similarly mottled back by a white or whitish nape band; this pattern is charac- teristic of /''//.so(/m (Fig. 7), melodns, pecuarius (Fig. 6), alexandrinus (Fig. 6), peronii {Fig. 6), venustus, mongolus, and rywntunus (Fig. 6); in rnodestus (Fig. 2&)2.x\A falklandicus {Y\g. 7) the nape is pale rufous. Within this group the ground color of the dorsal down varies from whitish to orange-buff to grayish brown. In the darker forms there is a tendency for dark bands to occur on the rump (wilsonia, falk- 14 JOSEPH R. JEHL, JR. Fig. 6. Chanidrius ulexandnnus, C. perotiii, C. pecnarni.s. and C. montanus. I(nidicus). occiput (icilsnuiii). wiriK (peronii, ulexitn- driiins), or midcrown (perouii, and perhaps no- vaeseelandiue); in jiecuarius there is a narrow, interrupted V-shaped fifjure on the bacl<. Within Chdnulrin.'i. however, the variations su- perimposed on the basic pattern are somewhat greater than in other wader genera. In liidficiihi and seniipahnatitu (Fig. 7) a conspicuous black band borders the nape band; in addition, two dark bands are present on the wing, and another extends from the h\p to the tail. C. duhius is similarly patterned, but in this species the entire crown patch is outlined in black. In t)i(lini()ps and rocif- erns (Fig. 7) the crown patch is also encircled b.v black down; in addition, there is a black midback stripe, a dark band from the hip to the tail, a bold ulnar stripe, and a black pectoral band. Chiir(idriii>i tricolluris and C. leschenmiltii are most similar to melcniopH and rociferuti but lack a black anterior border to the crown patch and possess a short, black midcrown stripe; leschei/ditltii lacks a pec- toral band. Two indistinct pectoral bands have been reported \nfs. In two species, bici)ict)is and iiordeseeldudiiie, the nape is lightly mottled and does not sharply separate the head from the back markings. A)iarhy)ichus: frontdJis* The chick of the Wrybill Plover is "pale grey speckled with black above, white below" (Oliver, 1955: 268); it apparently lacks a distinct dorsal pattern. Flii'l/iintis: nntchelli (1) According to Johnson (1965: 331), the Phegoniis chick is "uniform grey on the head and back with dove-grey throat, breast and ventral surface." In the specimen that I examined the dorsal down is brownish-gray mottled with black, the individual feathers being virtually identical to those of C7/r/- mdrius semip(th)idtus: the venter is grayish-white. There is no dorsal pattern. Pliiridli^: aprlairld (9), dottiinica (9), fiquatdrold (11), ob! Fig. 9. Bartra m ia longicauda, Nu nienius phaeapus, and Limosa haemastica. patch and extends onto the occiput. There are additional black markings on the shoulders and wings and lateral to the "diamond"; dark stripes from the hip lead to the tail. The only variation from this pattern occurs in A^. americanus, in which the borders of the crown patch are mottled, not solid (see illustration in Bent. 1929: PI. 20). Limosa: limosa (7), haemastica (13), lapponica (4), fedoa (6) The pattern of downy godwits (Fig. 2H ) is similar to that of curlews. It differs in having the crown patch solid, and the middorsal stripe, which is continuous with the "diamond" on the lower back, fails to bifurcate at the shoulder. As in the curlews, there are additional dark areas on the wings and back, and a line from the top of the thigh runs toward the tail. There is little variation in the basic pattern. In a few specimens of haemastica (Fig. 9) there is a second, smaller "diamond" on the rump. In all species except lapponica there is a prominent dark spot in the temporal region; in haemastica this mark is usually continuous with the crown patch. Chicks of limosa, lapponica, and fedoa are sim- ilar in color, the ground color of the down being light buffy brown; chicks of haemastica are much grayer and the dorsal markings are nearly black. Tringcu erythropus (10), totaniis (15), stagnatalis (1), nebrdaria (7), guttifer*, melanoleuca (1), flavipes (18), ocrophus (6), solitaria (1), glareola "(10) In this list of species I have placedflavipes next to melanoleuca. Nichols' (1923) arguments that these species are not closely related and that melano- leuca is a mimic oi flavipes. which may have been the basis for Peters' separation, are unconvincing. I have arbitrarily placed "Pseudototanus" guttifer next to nehularia. The genus Tringa comprises two groups, the "shanks" (erythropus through flavipes) and the "wood sandpipers" (ocrophus, solitaria, glareola), that are accorded generic status by the A. 0. U. (1957). If formal taxonomic recognition of these groups is deemed necessary, I would prefer to recognize them as the subgenera Totanus and Tringa, respectively. Chicks of the Nearctic species of the "shanks" (melanoleuca, flavipes; Figs. 10, 26,) are grayish with black markings dorsally, white ventrally; chicks of the Old World species average browner dorsally, but there is some overlap in coloration. The dorsal pattern (Fig. 21) consists of: 1) a solid crown patch, 2) three dark lines on the back that 18 JOSEPH R. JEHL, JR. U ^i^^. Fig. 10. Tringaflavipes. T. mehmoleuca. T. erythropus. and T. solitaria. may converge at the rump, and 3) dark lines from the hip to the tail that usually merge with the lateral stripes on the back. There are small dark patches on the wings and, in most individuals, on the lateral surface of the thigh. A dark line from the base of the bill to the anterior corner of the eye is present in all forms, and in all but nebidaria there is a prominent postorbital stripe that may extend to the occiput. Pattern variation is somewhat greater in Tringa, even within some species, than in most other genera. Some of the variations in head pattern are diagrammed in Figure 11. Variation in back pat- tern usually involves changes in the width or prominence of the stripes, but the lateral stripes may be interrupted, and in one specimen of T. totanus (UMMZ no. 64773) the middorsal stripe is broadened to a diamond with a light central area. Chicks of the "wood sandpipers" do not differ greatly in pattern from those of the "shanks," although the ground color of their dorsal down averages warmer brown. The middorsal stripe B D Fig. 11. Variation in tlie head patterns of tringine sandpipers: A. T. erythropus; B. T. melanokucaiC. T. fofanus, and D. T. ocrophus. RELATIONSHIPS IN THE CHARADRII 19 Fig. 12. Xeniis cinereus. Actitis macularia. and Heteroscelus incanus. continues onto the nape and is usually continuous with the posterior border of the crown patch. In glareola and ocrnphus a stripe from the base of the bill passes through the eye and ends on the occiput; in solitaria (Fig. 10) the preorbital portion of the stripe is absent and the postorbital portion is reduced to a small spot above the auriculars. Xenus: cinereus (1) Xenus chicks (Fig. 12) are light gray-brown with dark markings dorsally, whitish ventrally; the down on the neck and chest may be tipped with buff. The pattern (Fig. 2J) is like that of Tringa but the dark dorsal markings have been reduced, the crown patch to a single stripe from the base of the bill to the occiput, the three back stripes to a prominent middorsal line and two indistinct lateral lines; the lines from hip to rump are similarly reduced and prominent markings on the wings and thighs are lacking. As in Tringa there is a well- marked line through the eye. Actitis: hypoleucos (5), macularia (25) In Actitis the pattern is simple (Fig. 2K) and consists only of a single blackish middorsal stripe from the base of the bill to the rump; a conspicuous stripe from the base of the bill passes through the eye to the occiput. Chicks of both species are whitish below; dorsally those of hypoleucos tend to be buffy brown, those of macularia (Fig. 12) pale grayish, but there is some overlap. Catoptrophonts: semipahnatus (3) The ground color of Willet chicks varies from light grayish to buffy brown; the dark brown or blackish pattern is typical of Tringa. Heteroscelus: brevipes*, incmius (6) Tattler chicks (Fig. 12) are gray to bluish-gray above, white below; there is usually a faint tinge of gray on the throat and chest. Chicks that I have examined are nearly void of dorsal markings and have only a narrow transverse band of darker down on the rump. However, in Shortt's (1947) illustra- tion of incanus a faint central stripe is shown on both the back and head, and one of the chicks photographed by Weeden (1965) appears to have a crown stripe. Chicks of incanus have a well-marked line from the base of the bill to the anterior corner of the eye; in one specimen the line extends slightly 20 JOSEPH R. JEHL, JR. Fig. 13. Aphriza I'irgata, Arenarki interpres. and Arenaria melanocephala. posterior to the eye; chicks of brevipes are appar- ently identical to those of incanus, except that the eye stripe may be even more conspicuous. Subfamily ARENARIINAE Aphriza: virgata (1) I have examined only one partly-grown chick of this species (Fig. 13), but an e.xcellent photograph of newly hatched chicks has recently been pub- lished (Wetmore, 1965: 324). Both in color and pattern these chicks are nearly identical to those of Culidris canutus. The calidridine pattern is de- scribed below. Arenaria: interpres (9), melanocephala (6) Turnstone chicks are colored like those of Tringa totanus. Ridgway (1919: 51) described chicks oi A. interpres (Fig. 13) as: "Above light drab to drab- gray, irregularly mottled or marbled with black; forehead buffy drab-gray, with a median elongated black spot or irregular streak; a distinct loral streak of black, from bill to eye [which continues to the temporal region]; under parts dull white, the chest shaded with brownish gray." Chicks of A. melanocephala (Fig. 13) are similar, perhaps aver- aging slightly grayer. Except for the forehead stripe and the eye-line, there is no consistent pattern in this genus. Some chicks have two narrow, indistinct, parallel lines in the shoulder region and a mounted specimen in the Royal Ontario Museum (no. 35618) shows the typi- cal three-striped back pattern of Tringa. Subfamily SCOLOPACINAE Chicks of the Scolopacinae (except Scolopax ) and of the Calidridinae possess an unusual type of down, sometimes called "bottle-brush" feathers (Fig. 14). Barbs of these feathers are intertwined for most of their length, forming a rather rigid structure. From their tips short, white, densely packed barbules radiate and interlock with similar barbules of adjacent barbs, thus forming the (usu- ally) white puffs of dorsal down that characterize the downy plumage of most members of these subfamilies. Long, sparse, dark barbules originate laterally from the basal three-quarters of each barb. This portion of the feather corresponds to the dark "patterned" areas on other wader chicks. Therefore, the arrangement of the white dotting reflects a definite underlying pattern. In inter- preting the pattern diagrams for Gallinago. Lim no- drnmus, and Calidris one must recall that the areas diagrammed as dark will appear whitish on the specimens. Limnodromus: griseus (13), scolopaceus (3), semi- RELATIONSHIPS IN THE CHARADRII 21 Fig. 14. Structural differences in the superficially similar white-tipped dorsal down of; A. Calidris alpina: B. Thinocoms nimicivorus; C. Cursorius temminckii. For each species a single barb is illustrated. palmatus* The ground colorofdowitcher chicks varies from light golden brown {griseus; Fig. 15) to dark choco- late brown (scolopaceus and semipalmatus). Bottle-brush feathers are confined to the back and outline a definite pattern (Fig. 2L), which consists of paired lines on the midback flanked by lines of variable prominence on the sides of the body; short lines extend from the hip toward the tail. On the head, the crown patch is solid and is separated from the eye by a broad, whitish, supraorbital stripe; a continuous line extends from the base of the bill to the crown patch; malar stripes and dark throat patches are absent, although a small interramal spot may be present in a few individuals oi griseus. Coenocorypha: aucklandica (1) The single chick that I have examined is in such poor condition that little can be made of the pattern. A stripe from the base of the bill extends to the anterior canthus; another runs up the forehead and onto the crown, but its posterior limit cannot be determined. The chick is uniformly grayish brown above and below and is similar to chicks of Gallinago paraguaiae. Bottle-brush feathers are present but inconspicuous, and dark throat mark- ings are absent. Gallinago { = Capella; see Mayr 1963: 402): solitaria, hardwickii (1), nemoricola, stenura, megala*, nigripennis (1), macrodacttjla (1), media*, galli- nago (including delicata, 9), paraguaiae (1), no- bilis (1), undulata, stricklandii (1), jamesoni*, impenalis Fig. 15. Limnodromus griseus and Gallinago gallinago (dorsal and ventral). 22 JOSEPH R. JEHL, JR. Fig. 16. GalUnapo nobilis. G. strick-Iandii. and G. hardwickii. The linear placement of strick!(nidii and /w/k^- ridlis follows De Schauensee ( 1966: 98). Snipe chicks are not well represented in collec- tions and several species are undescribed. The species that I have examined are similar in color and pattern. Extensive dark brown or black mark- ings on a (usually ) chocolate ground color are found in most species, but in paraqnaiiie the black mark- ings are virtually absent, and in hardivickii and stricklandii (Fig. 16) the ground color is grayish or light tan. In all species a prominent black spot at the base of the upper mandible is separated from the crown cap by a light-colored transverse stripe from lore to lore; all species possess a malar stripe, which may be interrupted, and all but jamesoni show dark patches of variable prominence on the throat (Fig. 15); a small interramal spot is present in some species. The dorsal pattern of G. gallinago is shown in Figure 2M. There appears to be no regularity to the complex crown markings; the back pattern is like that oi LimnodromuH. Bottle-brush feathers occur in all snipe, being abundant in some (e.g., gallinago. Fig. 15), virtually absent in others (e.g., paraguaiue and hardwickii). De Schauensee (1966: 97) considered G. para- gudiae (in part) conspecific with G. gallinago but raised G. p. andina to species rank. I have not been able to examine a chick of G. paraguaiae (sensu De Schauensee), but Meinertzhagen (1926: 509) states that young of G. p. brasiliensis closely resemble those of G. g. gallinago. Young of G. p. andina differ markedly from those of G. gallinago and presum- ably from those of other races of G. parguaiae sensu Peters (1934); this supports De Schauensee's interpretations. De Schauensee (1966: 98) also sug- gested that G. jamesoni should be merged with G. stricklandii, but chicks of these forms differ slightly; in jamesoni throat patches are absent, whereas in stricklandii they are large and promi- nent. Lijni nocrijptes: minima* Jack Snipe chicks are nearly identical to those of Gallinago gallinago, but possess "a conspicuous and well-defined inframandibular bar of white" (Lowe, 1934:44). Scolopcur: rusticola (5), niira (see Vaurie, 1965), saturata, celebensis, rochusseyiii, minor (12) Only two species of woodcocks could be examined (Fig. 17); their back patterns are identical but the head markings of rusticola are less extensive than those diagrammed for minor (Fig. 2N). Presum- ably chicks of other species are similar; if variation exists it would most likely occur in rochussenii, the adults of which differ in pattern from other wood- RELATIONSHIPS IN THE CHARADRII 23 Fig. 17. Scolopax minor, andS. msticola. cock species. Young woodcocks are tan with dark brown mark- ings. A solid crown cap e.xtends from the base of the bill to the nape, where it broadens into a large dark patch. Three stripes on the back extend from the shoulder region to the tail, the central stripe being much broader than the lateral ones. There is a large dark area on the lateral surface of the thigh, and several small dark areas occur on the wings. The venter is unpatterned, but the neck and chest are more heavily tinged with brown than is the abdo- men. Unlike young snipe there are no dark mark- ings on the throat and bottle-brush feathers are not present. Subfamily CALIDRIDINAE ( = EROLIINAE of Peters, 1934) Calidns: canutus (3), tenuirostris*. alba (1), pusilla (40), mauri (6), ruficollis*, minuta (1), tem- minckii (2), subminuta, minutilla {21}, fuscicollis (4), bairdii{23), melanotos (6), acuminata* , mari- tima (3), ptilocnemis (13), alpina (22), ferru- ginea* ( = testacea) Eurynorhynchus: pygmeus (4) Limicola: falcinellus (6) Micropalama: Inmantopus (74) Tryngites: subruficoUis (1) Philomachus: pugncuc (3) Downy young of the six genera listed here (Figs. 18, 19) are so similar in pattern that they are most conveniently discussed together. This pat- tern, which is more easily diagrammed than de- scribed (Fig. 20) basically consists of irregular dark brown markings on the head and an "hour- glass" shaped figure on the back. As in the snipe, it is the bottle-brush feathers that give the clue to the pattern. The back pattern is not apparent in all chicks, although it is evident in C. ptilocnemis and M. himantoptts (Figs. 18, 19).Thismay bedueto the make of the skin, the age of the chick, or individual variation, but the age of the chick seems to be the most important variable. In older chicks the white barbules at the tips of the bottle-brush feathers are worn off or disentangled; thus, the barbs separate and the pattern becomes diffused. In spite of their great similarity, two groups of calidridine chicks can be recognized. In the first (pusilla, mauri, ruficollis, mi7uita. temminckii, minutilla, fuscicollis, melanotos, acuminata, ptilo- cnemis, alpina, ferruginea, pygmeus, falcinellus, himantopus, pugnax), the ground color of the dor- sal down is yellowish to chocolate brown; a rela- tively distinct dark crown patch is bordered by broad, whitish, supraorbital stripes that extend to the occiput; face markings are variable, but are usually not pronounced; all species have a varying amount of brownish wash on the neck and chest. In the second group (tenuirostris, canutus. alba, bairdii, maritinia, snbruficollis), the ground color of the dorsal down is brownish-gray and the dark markings are more extensive; the crown is inter- rupted by numerous white flecks and no definite cap is present; distinct supraorbital lines are ab- sent; dark face markings are common and more intricate than in the first group; all are whitish ventrally. Family RECURVIROSTRIDAE Subfamily IBIDORHYNCHINAE Ibidorhyncha: stru thersii* Ibisbill chicks are brownish-gray dorsally, pure white ventrally. Down on the nape is slightly paler than that on the back. The only evidence of dorsal pattern is an indistinct band of darker down that crosses the rump. A line from the base of the bill passes through the eye to the temporal region. Subfamily RECURVIROSTRINAE Himantopus: himantopus (7) Chicks of H. h. mexicanus, the only race that I have examined, are grayish above, white below. The most conspicuous features of the dorsal pat- tern (Fig. 2P) are the two dark lines that extend from the shoulder region toward the rump; in some 24 JOSEPH R. JEHL. JR. .7 •- -. Fig. 18. Calidris ptilocnemis, Eurynorhynchus pygmeus, and Limicolafalcinellus. Fig. 19. Micropalama himantopus, Tryngiies subrvficolUs, and Philomachuspugnax. RELATIONSHIPS IN THE CHARADRII 25 I Fig. 20. Himantopus himantopus (two left), and Recuroirostra americana. specimens these fuse on the midback and continue posteriorly as a single stripe; in others they extend to the rump before joining (see Fig. 20). Lines from each hip meet at the rump. A small spot occurs on the lateral surface of the thigh and a stripe on the wing parallels the radius. Head markings are variable; the crown and occiput are irregularly spotted; in all specimens there is a thin line from the crown to the nape, and in some there is an indistinct transverse band on the occiput. All chicks have a small spot at the anterior corner of the eye; the postorbital area is usually unmarked except for a spot at the temporal canthus, but one specimen has a well-marked postorbital stripe that reaches the occiput. The nape is pale gray and is not mottled. The pattern in most other races of himantopus is apparently like that of mexicanus. However, Housse (1945: 131) reported that young of niela- nurus have a black pectoral band. Chicks of no- vaezealandiae are "covered with a dark brown down" (Oliver, 1955:304). Cladorhynchus: leucocephala* According to Glauert and Jenkins (1931: 3), chicks of the Banded Stilt are "covered with whit- ish down black at the base." These authors did not note any dorsal pattern, and G. M. Storr, who reexamined two nestlings preserved in fluid, confirmed (pers. comm.) the lack of obvious pat- tern. Recurvirostra: avosetta (4), americana (17), novae- hollandiae*, andina* Avocet chicks (Fig. 20) are light tan with dark markings dorsally, white ventrally. Their dorsal pattern is like that of Himantopus, though less conspicuous. However, in avocets, the paired lines on the back always fuse at midback, never at the rump; in addition, a stripe from the base of the bill passes across the eye to the occiput. Family PHALAROPODIDAE Phalaropus: tricolor i6), lobatus (n),fi(licariiis (11) Phalarope chicks (Fig. 21) are yellowish with black markings dorsally, yellowish on the throat and neck, whitish on the belly. The ground color of the dorsal down varies; it is buff in tricolor, golden 26 JOSEPH R. JEHL, JR. Fig. 21. Phalaropus tricolor. P. lobatus. and P. fulicarius. yellow to light brown in lobipes, and light or dark brown infulicarius. The dorsal pattern is similar in all species. In lobipes (Fig. 2Q). for example, there is a dark crown patch, three stripes lead from the shoulder region to the rump, and a stripe extends from the hip toward the tail. All individuals pos- sess a small postorbital spot; some have a thin line from the base of the bill to the anterior corner of the eye. There is little pattern variation in this genus, although the prominence of the dorsal striping varies in lobipes and fitlicariiis, some individuals appearing almost black. All individuals of fitli- carius show an eye-stripe and a postorbital spot; these facial markings are absent in tricolor. The crown patch in tricolor is reduced to a thin central stripe that adjoins a narrow band on the occiput, thereby forming an anchor-shaped marking. Family DROMADIDAE Dromas: ardeola* Sharpe (1896: 29) described the young of the Crab-Plover as "covered with ashy-grey down above, slightly mottled with dusky; cheeks and under surface of body dingy grey, the general appearance of the bird being uniform." Family BURHINIDAE Burhinus: oedicnemus (5), senegallus*, vermicu- latus*, capensis*, bistriatus*, superciliaris, mag- nirostris (i) Esaciis: recnrvirostris (1), magnirostris Burhinus and Esacus chicks (Fig. 22) are grayish brown on the back and chest and dirty white on the throat and belly. The dorsal pattern (Fig. 2R) consists of two dark lines running nearly in parallel from the shoulder to the rump; a dark line from the hip extends toward the tail. There is an irregular dark patch on the forehead and scattered dark markings occur on the crown and occiput. Post- ocular stripes extend onto the occiput and continue down the neck; in some species these appear to be continuous with the paired back stripes. A promi- nent malar stripe leads from the base of the bill to theauriculars. Family GLAREOLIDAE Subfamily CURSORIINAE Pluvianus: aegypticus (1) The color pattern of the Egyptian Plover chick (Fig. 23) is plover-like. A broad, white nape band surrounded by black stripes that diverge from the posterior corner of the eye separates the lightly mottled, sand-colored back from the similarly col- ored crown. The venter is white, as are the wings, except for a narrow sandy stripe on the leading RELATIONSHIPS IN THE CHARADRII 27 h Fig. 22. Burkinus oedicnemus, and Esacus remrvirostris. edge. Cursorius: cursor (1), coromandelicus*, temrnin- ckii (2) Young of Cursorius fe m m i)ickii { Fig. 23 ) have the "upper parts mottled black, buff, and red-brown. The nape is crossed by a buff-white band. On the underparts chin and throat whitish, dull reddish brown on the breast" (Steyn, 1965: 32). Chicks of coromandelicus appear to be virtually identical to those of temminckii. but may lack a white nape (Ticehurst, 1926: 376). Cursorius cursor chicks are patterned like those of teynminckii, but are paler. Cursorius chicks appear to be covered with short, flat, contour feathers, but on close inspection it is evident that the "rachis" of each "contour feather" is but a single barb, and that each down feather contains si.x to eight such barbs. Barbules origina- ting from the barbs are stiff, closely packed, and lie in a single plane. Because of their structure, they do not merge with barbules of adjacent barbs and, thus, each unit looks and functions like an inde- pendent contour feather (Fig. 14). \ Fig. 23. Pluvianus aegypticus, Cursorius temminckii, and Gla- reola nuchalis. Fig. 24. Peltohyas australis. Rhinoptilus: africanus*, ductus, chalcoptenis* , bi- torquatus I know of no descriptions for young of Rhinop- tilus, but a chick of R. africanus (Fig. 26) photo- graphed by G. L. Maclean is much like that of Cursorius temminckii. It is reddish brown with patches of white and black scattered on the dorsum in no definite pattern; a fine black stripe extends from the base of the bill to the anterior corner of the eye; the venter appears reddish-brown but the color is not clearly visible in the photograph. Unlike C. temminckii, this species does not possess a white nape band. The curious flattened feathers noted in Cursorius also occur in R. africanus. 28 JOSEPH R. JEHL. JR. Fig. 25. Attagis gayi, A. malouinus. Thinoconis orbignyianus, and T. nimicivorus. The chick of Rhiiioptilus chalcopterus photo- graphed by Steyn (1966: 242) appears to be much more similar to chicks of Chamdriiis or Vanellus than to young of/?, africanus. Its dorsum appears uniformly colored, except for a broad, white, nape band that is bordered anteriorly by a black stripe. Unfortunately, neither the head pattern nor the feather structure are visible in Steyn's photograph. Peltohyas: austrulis* Chicks of this species (Fig. 24) are similar in color, pattern, and feather structure to those of Cursorius and Rhhwptilus africanus (description, in litt., and photographs by K. A. Hindwood). Peltohyas lacks a white nape band and is uniformly buffy on the venter. Subfamily GLAREOLINAE Stiltia: isabeUu* Glureola: pmtincola (1), maladivarum*, nord- manni, ocularis, nuchalis (1), lactea*. cinerea* Pratincole chicks are brownish-gray, indistinctly mottled with dark brown dorsally, pale brown or dirty white ventrally. There is no definite dorsal pattern. Apparently, most species have a fine line from the base of the bill to the eye. In G. nucfialis (Fig. 23) the line continues from the posterior corner of the eye to the occiput, where it meets a midcrown stripe; both of these markings are in- distinct. A midcrown stripe is also present in G. pnitincolu and mStiltia (Chaffer, 1949: PL 2). Family THINOCORIDAE Attagis: guyi (1), malouinus (1) Thinocorus: orbignyianus (1), rumicivonis (!) In A ttagis the buff or grayish-brown dorsal down is boldly crossed by wide dark markings that occur in no obvious pattern. The venter is uniformly grayish-brown in gayi. whitish with a buff tinge on the throat in malouinus. Chicks of Thinocorus are whitish ventrally, gray- brown flecked with whitish dorsally. In T. orbig- nyianus dark brown markings form a midcrown line and two short supraorbital lines, whereas in T. nimicivortis (Figs. 25, 26) the dark head markings are less distinct, and prominent white supraorbital lines are present. In each species there is a large dark V-shaped figure on the back, which in rumico- vonis is bordered laterally by a broad band of white-tipped feathers; in the single specimen of orbignyianus available this whitish band was ei- ther reduced or obscured by the make of the skin. In both species, but particularly in rumicivorus, white-tipped feathers are scattered over much of the dorsum. Although these feathers, at first glance, are reminiscent of "bottle-brush" feathers, their structure is quite different (Fig. 14). In this case the whitish, distal barbules are concentrated on one side of the barb, and tend to spiral along the RELATIONSHIPS IN THE CHARADRII 29 barb; in lateral view they look like wind-pruned spruce trees, with all the branches emanating from the leeward side. Seed-snipe chicks are shown in Figure 25. Family CHIONIDIDAE Chionis: alba (3), minor (1) Jones (1963: 67) described the young of C. alba as "clothed in a dark brown down which bears light- colored, hairlike tips and therefore presents a 'smutty brown' appearance." The natal down is followed by a second, dark gray down. The chick of C. minor is like that of alba, though perhaps slightly grayer; however, I have not been able to compare the chicks directly. There is no evidence of a distinct dorsal pattern in either species. RELATIONSHIPS IN THE CHARADRII In the following discussion relationships in- dicated by chick color patterns are compared with present interpretations of shorebird inter- relationships. In making these comparisons I have, unfortunately, been unable to consider fully the findings of Kozlova (1962) and Yudin (1965), be- cause translation of their works were unavailable. Where relationships indicated by downy plumages correspond to current views I have assumed that these are well founded; where they do not corre- spond, or where an alternative relationship is indicated, I have attempted to re-evaluate the evi- dence on which current views are based and to employ additional evidence in determining affinities. Within some families (e.g., Jacanidae, Haema- topodidae) chick patterns are nearly uniform; this precludes their use in suggesting relationships below the family level. The uniformity does confirm that species assigned to such families are correctly placed, but, more interestingly, it sug- gests that the patterns characterizing these groups may have remained essentially unchanged throughout much of the group's history. In other families (e.g., Scolopacidae) there is sufficient vari- ation that infrafamilial relationships can be post- ulated. Often the pattern characterizing one genus can be easily transformed into that of another, which I regard as evidence of close affinity; in some cases the reasons for pattern change can be infer- red. Where I have outlined possible pattern trans- formations, no phylogenetic implications are in- tended by the order of transformation given; the reverse transformation is equally possible. JACANIDAE, ROSTRATULIDAE The relative constancy of chick patterns in the Jacanidae precludes their use in clarifying inter- generic relationships. However, jacana plumages are so similar to those of young painted snipe that common ancestry may be postulated. Conceivably these resemblances might be convergent and in- dicate no more than similar selection for bold, disruptive patterns in wet tropical regions. How- ever, there are several reasons for rejecting con- vergence. 1) Nests are usually placed in different habitats, those of the jacanas on floating vegeta- tion, those of painted snipe in wet, grassy fields. Presumably, chicks are likely to be encountered near the nest; if so, it is hard to imagine a common background toward which the convergence might be directed. 2) The chick patterns in these families are distinctive and are unlike those of any other waders. Moreover, the patterns are comple.x, which minimizes the chance of independent origin and subsequent convergence. 3) The patterns of one family can easily be transformed into that of the other; by extending the postorbital stripes of the jacanas onto the back and adding a light center to the crown, and by losing a lateral stripe, a painted snipe pattern can be formed. The idea of relationship between these families is not new. It has been proposed by Furbringer (1888) and Beddard (1901). It is difficult to understand why Lowe (1931a) did not group these families together, since all but one of the osteological char- acters he presented for Rostratula are applicable to the Jacanidae. (The exception is the claimed pres- ence of supraoccipital foramina in Rostratula ben- ghalensis. I have not seen a skull of this species, but supraoccipital foramina are not present in the skull illustrated by Yudin (1965: 82), nor are they present in the one skull oi Nycticryphes that I have exam- ined.) Surprisingly, Lowe failed to note the sim- ilarity of the chicks. In groups as distinct as the Jacanidae and Ros- tratulidae, it is difficult to provide evidence that would prove close relationship. However, all the available evidence supports this idea and I know of none that contradicts it. I would place both families 30 JOSEPH R. JEHL. JR '^M Fig. 26. Left to right, top to bottom. Chicks of Irediparra galUnacea (Jacanidae): photo by Peter Roberts; Rostratula benghalensis (Rostratulidae): photo by K. A. Hindwood; Chanidrius modestus (Charadriidae): photo by Gordon Maclean; RhinoptUus ufricmms (Glareolidae): photo by Gordon Maclean; Tringa flavipes (Scolopacidae); Thinocorus nimicivorus (Thinocoridae); photo by Gordon Maclean. RELATIONSHIPS IN THE CHARADRII 31 in a single superfamily, Jacanoidea. DROMADIDAE This monotypic family is restricted to the Indian Ocean region. According to Lowe (1931b), its affinities are with the Glareolidae and Chionididae, but most other workers (e.g., Peters, 1934; Mayr and Amadon. 1951; Wetmore, 1960; Storer, 1960) place it near the Burhinidae. The similarity of adult Crab-Plovers in body form to the burhinids and in plumage pattern to the recurvirostrids might in- dicate that they are distantly related to that major wader line (see below). However, the young are unpatterned and cannot be used to support any relationship. Therefore, I retain the Crab-Plover in a monotypic superfamily, Dromadoidea. This species' unpatterned young are unusual, but its nesting habits and unpatterned eggs are unique in the Charadrii. Crab-Plovers nest in underground burrows; their chicks are apparently nidicolous and are fed in the burrows by the parents (Lowe, 1916b: 319; Archer and Godman, 1937: 497). Ticehurst (1923: 72) suggested that the absence of chick patterning in this species represented the retention of an ancestral condition, but this position seems untenable. Patterned chicks are of almost univer- sal occurrence in the Charadrii, and in those few- cases in which pattern is absent it can be most reasonably postulated that the loss is secondary. There are two possible hypotheses to explain pattern loss in the Crab-Plover. In nearly all hole- nesting species, the eggs and young are whitish and unpatterned; pattern loss in the Crab-Plover may represent a parallel response. On the other hand, since ancestors of the Crab-Plover were almost certainly surface-nesters with patterned eggs (a patterned egg has been reported: Anon., 1921: 21- 22) and young, it is possible that pattern was lost in response to selection for uniform coloration on a uniformly sandy substrate, and that burrow nest- ing evolved later. The fact that Crab-Plover young are countershaded (i.e., darker above than below), a condition that is presumably of value above ground, may support the latter interpretation. If pattern was lost subsequent to burrow-nesting, and if young remain in the burrows until the down feathers are replaced, it is curious that counter- shading has been retained. Lowe (1916b: 319-320) suggested that burrow nesting in Crab-Plovers may have evolved in re- sponse to predation on the eggs. However, I would suggest that this behavior may be an extreme condition derived from the habit of burying the eggs, ostensibly to protect them from intense solar radiation during the parents' inattentive periods. This function has been attributed to several spe- cies, including Charadrius alexandrius and C. pec- udriuti (Blaker, 1966: 100). Excessive heating may be a major cause of embryo mortality in desert birds (Maclean, 1967: 564). HAEMATOPODIDAE, RECURVIROSTRIDAE, BURHINIDAE Intrafamilial Relationships Within the Haematopodidae and Burhinidae the chick patterns are too uniform to indicate in- trafamilial relationships. In chicks of some mela- nistic oystercatchers (e.g., H. bachmani, H. o. unicolor) the normal, disruptive pattern is ob- scured by the darkened ground color of the dorsal down, but this doubtless reflects an over-riding secondary selection for cryptic coloration on dark nesting substrates, not common ancestry. In the Recurvirostridae the patterns confirm the obvious close relationship oi Himantopus and Re- curvirostm. Cladnrhynchus is similar to these gen- era and is intermediate in the characters that have been used to separate them (Ridgway, 1919: 436; Mathews, 1913-1914: 153); however, its chicks are pale and unpatterned. Since Cladorhijnchus nests on the shores of salt lakes (Serventy and Whittell, 1951: 158), where the substrate coloration may be unusually light, it seems reasonable to postulate that pattern and color have been lost secondarily as a means of enhancing cryptic coloration. Relationships within Himantopus and Recurvi- rostra cannot be inferred from the chicks, but variation in the chicks of H. himuntopus is greater than in any other species examined; this might indicate that more than one species of stilt should be recognized. The "absence" of pattern in H. h. novaezealandiae may be an artifact of description, since the pattern could be obfuscated by the dark down. On the other hand, the presence of a pectoral band in young of //. /(. melannrus, if Housse (1945) is correct, represents a decided departure from the usual pattern that deserves further investigation. Interfamilial Relationships The "two-lined" dorsal patterns in the Haema- topodidae, Recurvirostridae, and Burhinidae are extremely similar. The patterns are simple and could have arisen independently, but the families utilize such varied nesting habitats (e.g., rocky beaches, sandy beaches, dry grasslands, wet mar- 32 JOSEPH R. JEHL, JR. shes) that the independent evolution of similar disruptive patterns for a common background seems unlikely. Furthermore, close relationship between the Haematopodidae and Recurvirost- ridae is supported by osteological (Lowe, 1931b) and protein (Sibley, pers. comm.) evidence. Conder (1964: 77) has noted behavioral similarities be- tween these families, and Maclean (1966: 168) has pointed out that the curious "legs out" resting posture of these families is also shared by the Burhinidae. The Burhinidae were placed by Lowe (1931a) and Yudin (1965) nearest the Otididae (Gruiformes), but charadriine affinities are indicated by protein evidence (Sibley, 1960) as well as by chick plu- mages; thick-knee chicks differ markedly from those of bustards. The structure and patterning of individual midback feathers in the young of most species of Haematopodidae, Recurvirostridae, and Burhinidae are similar, being short, flattened, and crossed by a dark bar of variable prominence. The possible taxonomic significance of these feathers remains to be determined, for in species that nest in cooler or wetter areas the down is long and fluffy, and the barring is either reduced or obscured by the changed feather structure. Peters (1934), without explanation, placed the Ibisbill (Ibidorhynchu) in the Recurvirostridae. Other workers have placed it with the oystercat- chers (Seebohm, 1887), near the oystercatchers (Hartert, 1912-1921; von Boetticher, 1934), in the Vanellinae (includes also /f(?w«H^o/Ji(sandi?ec;