HBW15 - Family Text: Cardinalidae (Cardinals)

Class AVES


Suborder OSCINES


  • Small to medium-sized passerines with conical to very thick finch-like bill; male plumage coloration often brilliant, featuring pigmented reds and structural blues, females usually duller.

  • 11-24 cm.

  • America; introduced Hawaii.

  • Temperate and tropical forest, mixed woodland, savanna, arid scrub, suburban gardens; one species in grassland.

  • 11 genera, 42 species, 135 taxa.

  • No species threatened; none extinct since 1600.



Cardinalidae includes some of the most beautiful and best-loved birds of the Americas. The taxa traditionally placed in this family are, with the exception of the anomalousParkerthraustes andSaltator, a fairly homogeneous group of small to medium-sized finch-like birds ranging throughout much of North and South America. "Traditional" cardinalids have a more or less heavy, conical bill and, frequently, brilliant colours ranging from rich reds to glittering blues. They are almost entirely inhabitants of forest, forest edge and brushland, the chief exception being the grassland-specialized Dickcissel (Spiza americana), which looks like an emberizid sparrow.

The distribution of the family is centred on Middle America, its area of highest diversity ranging from the south-western United States southwards to South America north and west of the Andes. Other thanParkerthraustes andSaltator, which are probably not true cardinalids, only five of the twenty-five "traditional" species breed entirely outside these limits, although the Dickcissel and the Indigo Bunting (Passerina cyanea) are only marginal breeders within them. Of these twenty-five, the Rose-breasted Grosbeak (Pheucticus ludovicianus), Vermilion Cardinal (Cardinalis phoeniceus) and Yellow-green Grosbeak (Caryothraustes canadensis) have close relatives within the region. Only the Red-and-black (Periporphyrus erythromelas) and Glaucous-blue Grosbeaks (Cyanoloxia glaucocaerulea) represent genera absent from this core distribution, although both species have close, and possibly congeneric, relatives within it, the former being closely affiliated withRhodothraupis andCaryothraustes and the latter being close toCyanocompsa.

True cardinalids are entirely absent as breeding birds from the West Indies north of Trinidad and Tobago. The Rose-breasted Grosbeak winters sparingly in the western West Indies, and the Blue Grosbeak (Passerina caerulea), the Indigo Bunting and, rarely, the Painted Bunting (Passerina ciris) likewise spend the winter in that region. The Lesser Antillean Saltator (Saltator albicollis), however, is endemic on the islands of Martinique, Guadeloupe, St Vincent and Dominica.

Cardinalidae is one of the few nine-primaried oscine families with a fossil record predating human settlement. The Black-headed Grosbeak (Pheucticus melanocephalus) is known from the Pleistocene of Rancho La Brea, in California. D. W. Steadman and M. C. McKitrick assigned the proximal and distal ends of a right humerus and the distal end of a left humerus, collected in 1971 from a Pliocene (late Hemphillian) site in Chihuahua, in north Mexico, to the genusPasserina, although they noted that the specimens are "not adequate for a definitive generic assignment".

Since the late nineteenth century, cardinals have been considered to constitute a subfamily, variously named Richmondeninae, Pyrrhuloxiinae and Cardinalinae, or a tribe, Cardinalini, of a wider Emberizidae or Fringillidae, or they have, as in the present treatment, been regarded as a family in their own right.More crucial than debates about taxonomic rank, however, are the questions of how Cardinalidae is related to the other New World nine-primaried oscines, and of which species actually belong within the family. The two questions are interlinked, as studies comparing cardinalids with other nine-primaried groups have at times used, as examples of Cardinalidae, genera that probably do not belong within its ranks. For example, a 1995 molecular study of the Emberizidae (sensu lato) by M. Watada and colleagues, following C. E. Hellmayr, included two "cardinals" in the genusParoaria as examples of the Cardinalinae, notwithstanding the fact that in 1954 H. B. Tordoff had concluded, on the basis of skeletal morphology, that the genus was not cardinaline, and recent molecular analyses,including a 2002 study by T. Yuri and D. P. Mindell, support the placing ofParoaria among the tanagers (Thraupidae), although it is currently included within Emberizidae.

The species traditionally placed in Cardinalidae were generally put there because they were large-billed, often brightly coloured fruit-eaters and seed-eaters. Species having another bill morphology and different feeding habits were usually placed with other nine-primaried groups, tanagers in particular. Bill shape, however, has proven not to be a reliable phylogenetic character.

Hellmayr, in 1938, treated the group as a subfamily, Richmondeninae, of the Fringillidae. He included within its ranks the genera Saltator, Rhodothraupis, Caryothraustes, Periporphyrus, Pitylus, Gubernatrix, Paroaria, Richmondena, Pyrrhuloxia, Pheucticus, Hedymeles, Guiraca, Cyanocompsa, Cyanoloxia, Passerina, Porphyrospiza, Tiaris andSpiza, Hedymeles being used for the Rose-breasted and Black-headed Grosbeaks, now included inPheucticus. Tordoff removedGubernatrix, Paroaria, Tiaris andPorphyrospiza on the basis of their palatomaxillary structure, a decision supported by later molecular studies.

Some three decades later, in his subfamily Cardinalinae for J. L. Peters'sCheck-list of Birds of the World, R. A. Paynter included the following genera:Spiza, Pheucticus, Cardinalis, Caryothraustes, Rhodothraupis, Periporphyrus, Pitylus, Saltator andPasserina. His expanded genusPasserina incorporated five genera which had been treated separately by Hellmayr, namelyPasserina, Cyanocompsa, Cyanoloxia, Porphyrospiza andGuiraca. This treatment paralleled that used in 1888 by R. B. Sharpe, who includedCyanocompsa andCyanoloxia inGuiraca. It should be noted that, in the present arrangement, Pitylus is subsumed inSaltator, andPorphyrospiza, no longer considered cardinaline, is included within Emberizidae.

Twenty years after Paynter's classification, in 1990, C. G. Sibley and B. L. Monroe includedSpiza, Pheucticus, Cardinalis, Caryothraustes, Rhodothraupis, Periporphyrus, Pitylus, Saltator, Cyanoloxia, Cyanocompsa, Passerina andPorphyrospiza in a tribe Cardinalini within an expanded family Fringillidae. In 2007, J. Klicka and colleagues performed the first molecular analysis in which all of the cardinalid genera recognized by Sibley and Monroe were sampled. They examined 34 of the 42 species involved, and concluded that,although there is a valid, monophyletic cardinalid clade at the core of the Cardinalidae, the family as traditionally recognized was both polyphyletic, including genera and species that belonged elsewhere, and paraphyletic, excluding taxa that seemed properly to belong within its ranks. The results of this analysis expanded the membership of Cardinalidae to include not only the warbler-like "chats" of the genusGranatellus, an assignment first proposed in 2002 by I. J. Lovette and E. Bermingham, but also theAmaurospiza seedeaters and several genera of supposed tanagers, such asPiranga, Habia andChlorothraupis. Granatellus is currently placed with the New World warblers in the family Parulidae, whileAmaurospiza is usually included within Emberizidae. Klicka and colleagues found that bothCyanoloxia andAmaurospiza were nested withinCyanocompsa. In contrast,Parkerthraustes, formerly included within the genusCaryothraustes, andSaltator, the most speciose genus in the Cardinalidae as currently constituted, appeared not to be cardinalids at all.An additional genus,Porphyrospiza, often included in Cardinalidae, was associated instead withPhrygilus alaudinus, theBand-tailed Sierra-finch, a species usually placed in Emberizidae but probably a thraupid.

Klicka and colleagues identified five subclades within the newly reconstituted family. Two, theHabia-Chlorothraupis and theGranatellus subclades, contain only genera newly assigned to Cardinalidae and currently treated within, respectively, Thraupidae and Parulidae. The "traditional" cardinalids retained in the family fall into three subclades:


1) a "masked" subclade consisting ofPiranga, Cardinalis andCaryothraustes (includingRhodothraupis andPeriporphyrus, but excludingParkerthraustes);

2) a "blue" subclade consisting ofPasserina (includingGuiraca), Cyanoloxia (includingCyanocompsa andAmaurospiza) andSpiza; and

3) a grosbeak subclade containing thePheucticus species.


It should be noted that, in the present arrangement,Piranga is still treated under Thraupidae andAmaurospiza under Emberizidae.

Many of these conclusions, and those of earlier molecular studies, have already been accepted by the American Ornithologists' Union (AOU) and the South American Checklist Committee (SACC), although these bodies have not followed the study's proposed generic lumpings. Both have transferredChlorothraupis, Habia, Piranga, Granatellus andAmaurospiza to the family Cardinalidae and removedSaltator to "incertae sedis" near Thraupidae. The SACC has also transferredParkerthraustes, which does not occur in North America, from Cardinalidae to a position "incertae sedis".

This radical shift in classification, if correct-and the evidence supporting it appears to be strong-means that the present text, confined to species traditionally placed in the family, can give only a partial picture of the range of morphological and ecological variation within the cardinalid clade.To avoid confusion, the generaParkerthraustes andSaltator will generally be discussed only at the end of each introductory section.

Klicka and colleagues' 2007 study was not the first to question the traditional make-up of the Cardinalidae. An earlier molecular study by Klicka, K. P. Johnson and others, published in 2000, pointed to the misplacement ofPiranga outside the group and ofSaltator within it, and suggested that Cardinalini, as the authors classified it, was sister to the other four emberizid groups, namely Icterini, Emberizini, Parulini and Thraupini.

A. Grapputo and his colleagues, in a 2001 molecular study of relationships within the Emberizidae, found that the Cardinalidae, represented by the Northern Cardinal (Cardinalis cardinalis) and the Rose-breasted Grosbeak, clustered with theCalcarius longspurs and the snow buntings (Plectrophenax) as basal to other emberizids. These authors, however, did not examine other nine-primaried groups, and used the cardinalids essentially as an outgroup, and the support for this positioning was only moderate.

Placement ofthe generaPiranga, Habia andChlorothraupis within Cardinalidae was supported in 2002 by Yuri and Mindell, who found molecular evidence that cardinals and tanagers were sister-groups.J. Cracraft and F. K. Barker proposed that Cardinalidae and Thraupidae diverged between 11·6 million and 13·6 million years ago, during the Miocene. The cardinalid-thraupid clade, in turn, diverged from the emberizid lineage 14·1 million to 16·3 million years ago.

With its aberrant plumage and behaviour, the Dickcissel has for long seemed to be out of place within the Cardinalidae. W. J. Beecher removed it from the family altogether, placing it among the Icteridae, and R. J. Raikow, on the basis of appendicular mycology, considered the species to be intermediate between the emberizids and the icterids. Tordoff, however, found the palatal structure ofSpiza, with its unfused palatomaxillaries, to be typically "richmondenine" (cardinalid). Electrophoretic data obtained by J. W. Tamplin, J. W. Demastes and J. V. Remsen placedSpiza as the outgroup to all other cardinalids with the exception of Parkerthraustes, which is probably not a cardinalid, and Demastes and Remsen suggested thatSpiza might be closer toSaltator. The 2007 study by Klicka and colleagues, however, alliedSpiza with typical members of the "blue" subclade, such asPasserina, suggesting that its peculiar features may simply be adaptations to its grassland habitat.

The Yellow-shouldered Grosbeak (Parkerthraustes humeralis) was formerly included in the genuinely cardinalid genus Caryothraustes, although F. M. Chapman placed it inSaltator. It has long been recognized, however, that this species differs from its supposed congeners in morphology, plumage and behaviour. Tamplin and his colleagues, using data from starch-gel electrophoresis, suggested thathumeralis was not a cardinalid, and pointed out that its foraging behaviour, in which it is quiet, inconspicuous, and solitary or in pairs in mixed-species flocks, differed from the noisy, single-species flocking behaviour ofCaryothraustes. The 2007 Klicka-led study placed it among the tanagers, although both the AOU and the SACC retain it as "incertae sedis".

W. J. Bock, in 1964, supported the submerging ofRichmondena intoPyrrhuloxia, later to be renamedCardinalis, on the grounds that the morphological gap, manifested by a difference in bill shape, between the Pyrrhuloxia (Cardinalis sinuatus) and the Northern Cardinal is bridged by the third member of the genus, theVermilion Cardinal of northern South America (see Morphological Aspects). Bock suggested that the difference in bill shape between the two North American species might have evolved during a time when the two overlapped in habitat to a much greater degree than they do today.

Treatment of the generic boundaries within the Passerina-Guiraca-Cyanoloxia-Cyanocompsa complex has varied, some authorities lumping the genera in various combinations (see above) and others keeping them separate. L. L. Short considered the geographically widely separated Blue Bunting (Cyanocompsa parellina) and Ultramarine Grosbeak (Cyanocompsa brissonii) to be members of a superspecies, but Klicka and colleagues' 2007 study placedparellina either as a sister-species to the other members of their expanded Cyanocompsa, which included the Glaucous-blue Grosbeak and the emberizid Blue Seedeater (Amaurospiza concolor), or as one of the branches of a four-way polytomy including Cyanocompsa, Passerina andSpiza. The Ultramarine Grosbeak appeared to be most closely allied not to the Blue Bunting, but to the Glaucous-blue Grosbeak.

The status and affinities of the Blue Grosbeak, in particular, have been a point of disagreement for some time. Placed in the genusGuiraca by earlier authorities, such as Hellmayr, it has more recently been included inPasserina following Paynter's 1970 treatment. In a 1977 study of skeletal and morphological characters, J. J. Hellack and G. D. Schnell supported the maintaining ofGuiraca and allied it more closely toPheucticus, but this placement has not been supported by later studies. Klicka's molecular data suggest that the genusPasserina contains two lineages, a "blue" lineage consisting of the Indigo Bunting and Lazuli Bunting (Passerina amoena), and a "painted" lineage containing theRose-bellied Bunting (Passerina rositae), Orange-breasted Bunting (Passerina leclancherii), Varied Bunting (Passerina versicolor) and Painted Bunting, and that the Blue Grosbeak fits within the "blue" lineage.

There have been varying treatments of the members of the "Yellow Grosbeak complex", some recognizing them as separate species, the Yellow Grosbeak (Pheucticus chrysopeplus) of Mexico, the Black-thighed Grosbeak (Pheucticus tibialis) of Middle America and the Golden-bellied Grosbeak (Pheucticus chrysogaster) of north-western South America, and others treating them as conspecific. Short considered all three to form a superspecies with the South American Black-backed Grosbeak (Pheucticus aureoventris), noting areas of sympatry between the latter and the Golden-bellied Grosbeak and differences in the degree of sexual plumage dimorphism among members of the complex (see Morphological Aspects).

Paynter considered the Black-faced Grosbeak (Caryothraustes poliogaster) and the Yellow-green Grosbeak to be conspecific, but this idea has not been widely accepted, although the 2007 study by Klicka and others indicates that the two are sister-taxa. C. W. Thompson, in 1991, suggested that the eastern and western subspecies of the Painted Bunting should be regarded as separate phylogenetic species, on the grounds that there appeared to be little gene flow between them, but the suggestion has not been taken up.

The question of species limits among cardinalids has been particularly vexing for two species pairs, the Rose-breasted and Black-headed Grosbeaks and the Indigo and Lazuli Buntings, each with one representative in eastern and one in western North America. In each case, the members of the species pair hybridize in a narrow zone of overlap at or near the western edge of the Great Plains. Presumably, eastern and western populations became isolated during the Pleistocene, but came into secondary contact at some point following the retreat of the glaciers. For the grosbeaks, especially, this situation has raised questions as to whether one or two species should be recognized, and has also, more productively, led to a number of studies on the nature and effect of evolutionary isolating mechanisms and the viability of hybrids.

The two grosbeaks have generally been maintained as separate species, although D. A. West and, later, A. R. Phillips favoured combining them. In the United States, R. L. Kroodsma found few hybrids in North Dakota, where there is little suitable habitat in what might otherwise be an area of range overlap. He demonstrated that male grosbeaks in areas of allopatry in North Dakota were more likely to attack mounted specimens of their own type than they were those of the other member of the species pair, suggesting that males, at least, discriminate between the two species on the basis of plumage. In South Dakota, B. W. Anderson and R. J. Daugherty found that Black-headed and Rose-breasted Grosbeaks hybridized within a zone some 84 km in width along the Missouri River. Even within this range, where hybrids might be expected to predominate, parental phenotypes occur in numbers too high to be explained by genetic recombination or immigration. Anderson and Daugherty found evidence that female Black-headed Grosbeaks, at least, prefer to mate with males of their own morphotype. In addition, the clutch size of hybrid pairs was lower than that of "pure" Rose-breasted or Black-headed pairs, suggesting that hybrids have reduced viability. On the basis of this evidence, the authors concluded that the two forms should be regarded as separate semi-species within a superspecies. A 2009 molecular study by R. D. Mettler and G. M. Spellman re-evaluated Anderson and Daugherty's work and suggested that hybridization was less frequent than had been estimated on the basis of phenotype alone. These authors concluded that the hybrid zone between the two species was stable, and supported the conclusion that hybrids within it were less fit than their parents.

The question of the relationship between the Indigo and Lazuli Buntings has been confused by Klicka's cytochromeb analysis ofPasserina, which suggested that the two were not, in fact, sister-species, the Lazuli Bunting being more closely related to the Blue Grosbeak. A later study by M. D. Carling and R. T. Brumfield, using ten nuclear loci, did, however, support the traditional sister relationship between Lazuli and Indigo Buntings. Carling, I. J. Lovette and Brumfield have recently estimated that the two species diverged about 1·04 million years ago, during the middle Pleistocene.

Calculations based on the presumed dates for the expansion of oak (Quercus) savannas in central North America suggest that Indigo and Lazuli Buntings may have been in secondary contact for at least 3500 years, and they have certainly been so for the past 120 years. Their ranges overlap within a contact zone that may be as little as 40 km wide, although Sibley and Short estimated that Indigo and Lazuli Buntings overlapped in Nebraska across a zone more than 400 miles (640 km) across containing both hybrid and "pure" forms. This difference may partly reflect the way in which introgression has been measured in different studies: phenotypically, for example, as opposed to the findings of molecular-genetic analyses.

Studies of bunting hybridization have produced conflicting results, morphological analyses suggesting a high degree of interbreeding, but field studies finding that thereal frequency of hybridization was much lower. Because different alleles may introgress at different rates, the presence of hybrid gene combinations may occur over a narrower or broader geographical distance, depending on which alleles are involved. Hybrid pairs, especially those in which the female parent is herself a hybrid, exhibit lower reproductive success than do single-species matings, suggesting that there is selective pressure against hybridization. This pressure may be the reason why the zone of overlap is so narrow. Female mate selection appears to strengthen the isolating mechanisms between the two species. Within their zone of overlap, male Indigo and Lazuli Buntings act in some ways like members of the same species; they defend territories against each other, and they respond aggressively to each other's songs. Females, on the other hand, show a preference for mating with their "own" plumage types, or with males whose song phrases match those of their own plumage type. Interestingly, hybrid females mate randomly on the basis of plumage, but prefer Lazuli song types.

M. C. Baker and J. T. Boylan concluded, after a four-year study of colour-ringed individuals, that Indigo and Lazuli Buntings should continue to be regarded as separate species. Males of each species, however, can learn one another's song traits, and this learnt behaviour can contribute to interspecific mating and introgression, despite the genetic differences between the two species.

Recent work suggests thatthe measuring of mitochondrial genetic distances between closely related species pairs ("DNA barcoding") can provide a means by which to assign specimens to species, pairs with hybridizing species tending to have smaller distances than pairs with non-hybridizing species. Distance data for Indigo and Lazuli Buntings (taken outside the hybrid zone) are higher than average, but this may not be a measure of overall genetic distinctiveness.

Kroodsma speculated in 1975 that introgression in North Dakota and eastern Montana, in the north United States, was relatively high, but S. T. Emlen, J. D. Rising and W. L. Thompson, who studied populations farther south, in Nebraska, suggested that it was minimal in either direction. Carling and Brumfield's later studies, which benefited from technology that allowed detailed studies of individual loci, found that the degree and direction of introgression varied for different alleles. Mitochondrial haplotypes (mtDNA) showed greater introgression fromcyanea (Indigo Bunting) intoamoena (Lazuli Bunting), but alleles from two loci located on the Z chromosome showed much greater introgression in the opposite direction. In general, there was greater introgression among alleles of z-linked loci fromamoena intocyanea than the reverse, suggesting that z-linkedamoena alleles may be selected against more strongly when present in a cyanea-type hybrid than the reverse.

Carling and Brumfield also found that both mitochondrial haplotypes and z-linked alleles introgressed less thandid autosomal alleles. Because mitochondrial haplotypes are passed on only by females, and because females are the heterogamic sex in birds, this finding supports Haldane's Rule, which predicts that, in hybrid populations, if one sex is absent, rare or sterile it will be the heterogamic sex. In this case, female hybrid buntings should be, and apparently are, less fit than males.

By determining which alleles show the least degree of introgression-in other words, the ones most resistant to being taken up in the genomes of the other species-it may be possible to identify the specific genes involved in reproductive isolation. Carling and Brumfield found that a single locus on the Z chromosome, VLDR9, showed introgression across a band only 2·8 km wide. Mutations of the VLDR gene, which is involved in yolk-precursor formation, have been found to lower the egg-laying ability of some chickens, and it is possible that a similar effect may be involved in both the initial generation and the current maintenance of reproductive isolation between Indigo and Lazuli Buntings, despite the probability that some level of gene flow between the two has been going on for some time, even while the two species were in the process of speciation.

The systematics of the large genusSaltator, and its position within the thraupid complex to which it probably belongs, remain unclear. The genus may be polyphyletic. Hellack and Schnell, in their 1977 study, concluded that the Rufous-bellied (Saltator rufiventris), Streaked (Saltator striatipectus), Masked (Saltator cinctus), Black-throated (Saltator atricollis), Golden-billed (Saltator aurantiirostris) and Orinocan Saltators (Saltator orenocensis) bore little similarity to the other species, and recommended that the last three be removed from the genus. Further, J. P. O'Neill and T. S Schulenberg have emphasized that the Masked Saltator, with its unique colour pattern and particularly deep bill, has no obvious close relatives. One should note, incidentally, that Hellack and Schnell treated the Streaked Saltator as conspecific with the Lesser Antillean Saltator, referring to it under the latter's scientific name ofS. albicollis; it has since become evident, frommitochondrial-DNA studies of Antillean birds and specimens from Peru and Panama, that the two are separate species (see below).

Some 30 years later, in 2007, Klicka and co-workers examined ten species taxa, namelygrossus, coerulescens, striatipectus, atripennis, atriceps, maximus, nigriceps, aurantiirostris, atricollis andrufiventris, and concluded that all except the last-named were properly placed in Saltator. They found thatrufiventris was nested instead with the thraupid genera Delothraupis andDubusia. A new generic name may have to be found for this species.The Slate-colored (Saltator grossus) and Black-throated Grosbeaks (Saltator fuliginosus), allopatric species sometimes treated as conspecific, were formerly placed in a separate genus,Pitylus, on the basis of their bill and tarsal morphology; they differ also from all except the Thick-billed Saltator (Saltator maxillosus) in being sexually dimorphic. Molecular data, however, reveal that these species are nested withinSaltator, and the 2007 Klicka-led study foundSaltator nigriceps, the Black-cowled Saltator, to be basal to bothaurantiirostris andgrossus, thoughfuliginosus was not examined.

G. Seutin and colleagues, in 1993, separated the Lesser Antillean Saltatorof the West Indies from the continental Streaked Saltator on the basis of data indicating that the island populations differ genetically from those of Peru and Panama to an extent greater than that found between many other congeneric passerine species. Even within the Streaked Saltator itself, however, genetic differences are considerable, populations in Peru also differing from Panamanian individuals at levels consistent with congeneric species differentiation. This suggests thatS. striatipectus may have to be split further.

The question of species limitsin Saltator has been vexed by lack of information, including an absence of data on zones of overlap and/or intergradation between closely related forms. The Golden-billed Saltator, for example, has been considered conspecific with both the Black-cowled Saltator (but see above) and the Thick-billed Saltator; Short treated all three as members of a superspecies, but noted that the Golden-billed and Thick-billed Saltators interbred extensively in Corrientes Province, in north-east Argentina. The Black-cowled differs, however, from the Golden-billed Saltator in vocalizations, and there seems to be little or no intergradation between the Thick-billed Saltator and the subspecies parkesi of the Golden-billed Saltator in Rio Grande do Sul, in south-east Brazil, where the ranges of the two meet. Short considered that the Green-winged Saltator (Saltator similis) and the Greyish Saltator (Saltator coerulescens) formed a superspecies, although they are sympatric in some areas.

Finally, the major 2007 study undertaken by Klicka and colleagues found that the Many-colored Chaco Finch (Saltatricula multicolor), usually treated as a member of the family Emberizidae, was nested withinSaltator, in a position close to the Black-throated Saltator. Although the SACC has declined to include the Many-colored Chaco Finch withinSaltator, it may consider the possibility of transferring the Black-throated Saltator toSaltatricula in the future.

The mysterious "Townsend's Bunting", known from a single specimen received by J. J. Audubon in 1833, and described asEmberiza townsendi, probably represents a Dickcissel lacking carotenoid pigments in its plumage.


Morphological Aspects

Typical cardinalids are rather stocky, medium-sized to small finch-like birds witha heavy bill and a relatively short tail,Cardinalis, the only crested genus in the family, being somewhat longer-tailed. The males tend to be slightly larger than the females. Dickcissel males are significantly heavier than females for most of the year, the difference being due to increased amounts of protein and water. Females retain their stores of vernal fat for longer than do males, reducing the difference between the sexes in summer.

Northern Cardinals are non-migratory (see Movements), but are capable of adjusting their metabolism to deal with the stresses of harsh north-temperate winters. Thermal regulation in winter appears to be more energetically costly for cardinals at the northern end of their range than is breeding. Wintering cardinals studied by C. E. Sgueo in southern Ohio increased their daily energy expenditure, raised their metabolic rate and took on extra fat, increasing their average body mass from 41·5 g in summer to 47·2 g in winter. Pectoral-muscle mass did not increase.

In the case of the Painted Bunting, skull pneumatization takes an estimated 46 days to complete. In the Black-faced Grosbeak, skull pneumatization is completed posteriorly, rather than anteriorly, a pattern seen in a number of tropical species.

Bill shape varies considerably even within genera of Cardinalidae. WithinCardinalis, the maxilla, or upper mandible, of the Pyrrhuloxia is remarkably like that of a parrot (Psittacidae), being short, thick and deep, with a strongly curved culmen and a strongly notched tomium; the species' former generic name ofPyrrhuloxia alludes to its supposed similarity to the fringillid generaPyrrhula andLoxia. The maxilla of the Northern Cardinal is longer and less curved, with little or no tomial notch. The lower mandibles of the two species differ even more, that of the Pyrrhuloxia being deeper and having a ventral bony boss at the gonys. The bill of the Vermilion Cardinal is intermediate between those of its two congeners. Jaw musculature in the genus parallels the differences in skull and bill morphology, the longer-billed Northern Cardinal having generally larger jaw muscles that allow it to crack larger seeds, which can be held farther forward in the bill for cracking than is the case with the Pyrrhuloxia.

Northern Cardinals in Bermuda appearto have undergone shifts in bill shape since their introduction in the eighteenth or nineteenth century (see Relationship with Man). In 1901, O. Bangs and T. S. Bradlee noted that the bills of this species from Bermuda had a grooved upper mandible, a feature absent from mainland populations of Northern Cardinals but, oddly enough, present in the Vermilion Cardinal. K. L. Crowell found that Bermuda cardinals had broader, deeper and slightly shorter bills than those of mainland individuals, possibly an adaptive response to reduced competition from the limited landbird fauna of the islands.

In common with other nine-primaried oscines and fringillids, but unlike most other passerines, cardinalids produce substantial amounts of 3-methyl fatty acids in the feather waxes secreted from their uropygial glands. Unusually, and perhaps uniquely, among nine-primaried oscines, the Dickcissel produces feather waxes that are low in 3-methyl but high in 2-methyl fatty acids, the type of compounds common in most other passerine groups.

Excluding the generaParkerthraustes andSaltator, almost all cardinalids are sexually dichromatic in plumage, often strikingly so, with bright colours largely restricted to the males. The two species ofCaryothraustes, in which the sexes are similar, are the only exceptions. The plumage patterns of male and female Black-thighed Grosbeaks are essentially similar, females differing from males only in being slightly paler on the head and breast, with more olive scaling on the back and, usually, less white in the wing. Bright male colours include various shades of red or pink, particularly in members of the "red" subclade, but also in somePasserina species, yellow, found in NeotropicalPheucticus and the Orange-breasted Bunting, and an array of structural blues in the "blue" subclade.

In some cases, females may develop male characteristics, such as the rose-red wing-linings found on some female Rose-breasted Grosbeaks. One Northern Cardinal ringed in Tennessee, in the eastern United States, was a bilateral gynandromorph, having its right side bright red and its left side greyish-olive. Several researchers have remarked that the strong sexual dimorphism exhibited by the Northern Cardinal, a non-migratory monogamous species in which both parents care for the young, is an exception to the general tendency for such birds to be monomorphic.

Ornamental plumage is the rule, rather than the exception, among cardinalid males. Except for the crests ofCardinalis, however, feather ornamentation is a matter of colour, rather than of modified feather shape. Ornamental colours may signal the health or condition of their possessor, but crest length of the Northern Cardinal appears not to do so. Instead, it probably functions as a signal of behavioural intent (see General Habits). Male Blue Grosbeaks sometimes elevate their crown feathers, giving the appearance of a crest, and a male Crimson-collared Grosbeak (Rhodothraupis celaeno) observed in Texas raised its crown feathers when singing, lending it a crested appearance and a cardinal-like silhouette.

Red and, presumably, yellow coloration in cardinalids is produced largely, though not entirely, by carotenoid pigments. J. Hudon has identified canthaxanthin, astaxanthin, phoenicoxanthin, doradexanthin andε,ε-carotene-3,3'-dione among the red carotenoid pigments (4-keto-carotenoids) of the Northern Cardinal, and astaxanthin and canthaxanthin among the rose-pink breast pigments of the Rose-breasted Grosbeak. Northern Cardinals also exhibit minor amounts of rhodoxanthin, a xanthophyll.

Birds are unable to synthesize or store carotenoids, but must take them up in their diets. S. U. Linville and R. Breitwisch found that a reduction in carotenoid-bearing foods, in this case a failure of grape and other fruit crops, resulted in a slight, but significant, reduction in colour brilliance among male cardinals in Ohio. Birds can, however, transform carotenoid precursors which they obtain in their diet into different forms of pigment deposited in their feathers. Northern Cardinals derive their red pigments primarily from orange or yellow carotenoid precursors; males can grow pale red feathers even if they are fed only with yellow carotenoids.

A single maleNorthern Cardinal collected in Louisiana, and studied by K. J. McGraw and his colleagues, had only yellow, instead of red, pigments. Presumably, this was the result of a genetic mutation that interrupted the bird's ability to complete the metabolic pathways necessary to produce the compounds responsible for normal coloration. Interestingly, a similar mutation appears to be responsible for the yellow pigments in the Western Tanager (Piranga ludoviciana), a species now known to be a cardinalid (see Systematics) with close, red-plumaged relatives.

The presence of carotenoids can be an indicator of a bird's condition, at least at the time when its feathers are growing. In the case of Northern Cardinals, carotenoid feather ornaments act as guides to the health and fitness of prospective mates, or serve to reinforce existing pair-bonds, for both male and female birds. Males display the bright red plumage of the upper breast to females, while females expose their red underwing-coverts to males (see Breeding). Red coloration seems not to serve as a status signal for cardinals outside the breeding season; for example, L. LaReesa Wolfenbarger found that the artificial intensifying of the birds' red colour gave them no advantage over other males at a food source.

Brown colours, such as those in the wingbars of the Blue Grosbeak, result from a mix of eumelanins and phaeomelanins.Melanin ornaments, such as the black face mask of the Northern Cardinal, are often associated with high testosterone levels and increased aggression.In a study by J. M. Jawor and her colleagues, female cardinals having a larger, darker black face mask were found to be more aggressive towards mounted models, but were also better provisioners of their young. Males with a larger face mask, however, had lower reproductive success, perhaps because they spent more time in aggressive behaviour and less in provisioning their broods.

The blue colours in the feathers ofPasserina, Guiraca, Cyanoloxia andCyanocompsa are, like almost all such colours in birds, structural, being produced by light scattering from the spongy medullary layer of feather barbs, rather than by pigment. Males ofPasserina andCyanocompsa species employ structural colours, pigments including eumelanins, phaeomelanins and carotenoids, and combinations of the two to produce a range of colours, visible to birds, across a greater spectrum than the human eye can perceive.

Plumage patternsof Cyanocompsa are almost entirely the result of blackish eumelanin-pigmented patches and structural ultraviolet/blue colours that, together, appear to human eyes as deep blue. The ultraviolet-rich blue and turquoise plumage of the Indigo Bunting is entirely structural. Blue Grosbeaks and Lazuli Buntings add deep red-brown patches produced by phaeomelanin. The belly-patch colour of the Lazuli Bunting, appearing to humans as white, is structural, resulting from the morphology of the species' unpigmented feathers. The reds, yellows and oranges of Painted, Varied, Rose-bellied and Orange-breasted Buntings are carotenoid colours, while the Painted Bunting's green back, the purple hues of the Varied Bunting and the Rose-bellied Bunting's pink belly are apparently produced by a combination of carotenoids and structural colours.

M. C. Stoddard and R. O. Prum concluded that blue structural colours evolved only once in the "blue" subclade (see Systematics), and that phaeomelanin-based and carotenoid-based colours in subclade-members were later evolutionary developments; the authors did not examineCyanoloxia orAmaurospiza. The ancestral plumage probably resembled the deep blue coloration of Cyanocompsa, the colours increasing in range and brilliance during the course of the group's evolution. Loss of brilliance in the colours of the Lazuli and Varied Buntings probably represents later, secondary developments.

In addition to white, black, violet, blue, turquoise, green, yellow, orange, brown, pink and red, the feathers ofPasserina andCyanocompsa display ultraviolet (UV), UV-yellow and UV-red, the last two being colours visible to us as yellow and red, but reflecting strongly in the ultraviolet as well. Avian eyes contain four types of cone receptor, with one sensitive to UV-violet wavelengths in addition to those responsive to blue, yellow and red, and the evolutionary significance of plumage colours may be driven by the way in which the birds themselves perceive them. The breast of an Orange-breasted Bunting, which reflects strongly in the ultraviolet, in addition to being pigmented with yellow carotenoids, must appear an imaginable shade to a bird's eye.

Feathers on the Blue Grosbeak reflect more strongly in the ultraviolet than in the visible blue spectrum. A. J. Keyser and G. E. Hill found that males showing little variation in their visible blue colours varied in the intensity with which they reflected ultraviolet wavelengths. The richness of structural blue and UV colours may be affected by the rate at which the feathers grow, which is, in turn, a reflection of nutritional condition during the autumn moult. The degree of brightness of Lazuli Buntings, for example, may be due more to environmental factors such as moult than to direct genetic control. Keyser and Hill found that male Blue Grosbeaks that grew their tail feathers at a faster rate also had brighter rump and breast plumage, suggesting that structural brightness may be related to the individual's condition.

In another study, A. S. Barreira and her colleagues found differences in colour intensity and UV reflectance between subspecies of the Ultramarine Grosbeak occupying different habitats. Males of the subspeciesargentina, which lives in semi-open areas, had brighter and less saturated coloration, with higher hue and lower UV reflectance, than did males of the racesterea, which occupies densely vegetated forests where luminosity is lower and blue wavelengths prevail. The differences could be a response to the quality of ambient light in the two habitats.

Moult strategies of cardinalids resemble those of many other songbird groups, although, as with numerous aspects of their biology, little is known of the situation with regard to tropical and south-temperate species. Juveniles moult at least once in their first year, this resulting in a plumage that is replaced, in subsequent cycles, by the adult plumage. Depending on the species, this adult plumage may be much the same in appearance throughout the year, being replenished by an annual moult, or there may be two moults per year, producing an alternation of breeding and non-breeding plumages. In the case of the Orange-breasted Bunting, these two plumages are essentially identical, and brightly coloured.

Species ofCardinalis, Passerina andCyanocompsa, and possibly the Rose-breasted Grosbeak, in the genusPheucticus, have been described as having an extra juvenile moult just after fledging and before the moult producing the first-year plumage. For individuals that hatched late in the season, these moults may overlap. In the case of young Rose-breasted and Black-headed Grosbeaks, the moult producing first-year plumage may overlap with that producing the first breeding plumage.

S. Rohwer and others have claimed thatPasserina buntings undergo a further moult, this one involving the rectrices, the three outermost primaries and, typically, the three innermost secondaries and all of the body feathers. With Indigo Buntings this produces the first sexually dimorphic plumage, young females then resembling adult females and young males resembling males in non-breeding plumage. Indigo Buntings generally undergo this moult on the wintering grounds; fewer than 10% of individuals complete it prior to autumn migration. Moults during the first year may not be clear-cut events, however, and R. B. Payne has argued, on the basis of his studies of Indigo Buntings, that this pattern in fact represents an extended overlap between the juvenile and first-year moults, rather than a separate moult.

According to a study by T. C. Grubb and his colleagues, the rate of feather growth during moult in Northern Cardinals is affected by ambient temperature, as well as by health and condition. Rectrix growth is slowest during the winter. It is more rapid in males than in females, and in adults than in juveniles.

Northern Cardinalsmoult their wing and tail feathers simultaneously. Early-hatched young of this species often undergo a complete moult, but the moult of the primary feathers, which may begin after the onset of moult of the wing-coverts, becomes less extensive among individuals hatched later in the season. These late-hatched cardinals may retain their flight-feathers during the winter. In Kentucky, juvenile Northern Cardinals stop moulting at some time in November, regardless of when they began or whether the moult is complete. Those moulting into first-year plumage usually replace their wing feathers in typical passerine fashion, starting with the innermost primaries first, but retaining some outer primaries from the juvenile plumage. This is the same sequence as that followed by adult Pyrrhuloxias, whereas young Pyrrhuloxias begin the moult from a middle primary. The reason for this marked difference is unknown.

Northern Cardinals and Pyrrhuloxias moult once per year. Breeding plumage is gained not by moult, but by the wearing-away of the grey or brownish tips of the body feathers. Other North American species usually moult twice annually. The pre-breeding moult, however, may be very limited or even absent, particularly in the genusPasserina, in which it may be restricted to the feathers of the head, particularly those around the auriculars, eyes, bill and throat. Other changes into breeding plumage are the result of feather wear. Feather wear may be responsible also for the gradual decrease in UV chroma and hue in male Ultramarine Grosbeaks after the conclusion of the breeding season.

Before the onset of the breeding season,the process of abrasion wears away the buff tips of the male Blue Grosbeak's otherwise blue body feathers and the brown edgings of the body feathers of the Varied Bunting. In adult male Lazuli Buntings, the change from dull winter plumage to bright breeding dress is accomplished mostly by abrasion of the buff feather tips, but the Indigo Bunting changes to breeding plumage through an extensive moult. Second-year male Lazuli Buntings are similar in plumage to the adults, but they remain dull because, in contrast to the situation with adults, the feather tips generally do not wear away. Why this is so-whether it is the result of a hormonal difference between younger and older males, or whether there is some difference inbehaviour that affects feather wear-is not understood.

Among migratory species, the post-breeding moult is generally completed on the breeding grounds or, sometimes, as may be the case with some Rose-breasted Grosbeaks, for example, it is begun on the breeding grounds and completed after migration. Most individuals moult into breeding plumage on their wintering grounds, but some Black-headed Grosbeaks may complete this moult during the course of the northward migration, and some Indigo Buntings arrive on their breeding grounds while still growing blue feathers on the head.

Some Black-headed Grosbeaks begin, or even complete, the post-breeding moult on the breeding grounds. Most, however, undertake moult migrations after the breeding season, transferring to monsoonal staging areas in the south-western Unites States or north-west Mexico, where the moult takes place. This strategy allows these birds to leave the nesting areas, where food may be scarce during late-summer dry spells, and take advantage of areas where a food supply sufficient to fulfil the energy needs of moulting birds is still available.

Blue Grosbeaks, Lazuli Buntings and western populations ofthe Painted Bunting are also known to be moult migrants. Some Lazuli Buntings fly directly to their wintering grounds in western Mexico and moult there. Most, however, begin their moult on the breeding grounds, interrupt it for migration, and complete it either in southern Baja California or in an area reaching from southern Arizona and south-west New Mexico south into northern Sonora, in extreme north Mexico. Here, in these "migratory hot spots", they remain for approximately one month while the intensity of moult increases to about twice that on the breeding grounds. Indigo Buntings are not moult migrants, and it is unclear where Indigo x Lazuli Bunting hybrids undergo their post-breeding moult.

Delayed plumage maturation, in which juvenile males take two years or longer to gain full adult colours, is recorded for the Yellow, Rose-breasted, Black-headed and Blue Grosbeaks and for the Indigo, Lazuli, Varied, Painted and Orange-breasted Buntings. Second-year male Black-headed Grosbeaks vary in plumage, ranging from individuals that resemble adult females to others that are much like older males. The difference between first-year and adult male Indigo Buntings is poorly marked. Of particular interest is the fact that Orange-breasted and, apparently, Rose-bellied Buntings, unlike other Passerina species, exhibit delayed plumage maturation also of females, an unusual condition among passerines. Females of these two species retain a duller juvenile plumage until after their first breeding season.

Hill found that delayed plumage maturationamong Black-headed Grosbeaks was correlated with lower testis mass in yearling birds. The small likelihood that a yearling male grosbeak will be able to win a high-quality territory and breed successfully, coupled with the need to avoid aggression from older males (see Breeding), may have favoured the evolution of delayed plumage maturation in this species.

Yearling cardinalids are, nevertheless, physically capable of breeding. After the first week on the breeding grounds, the sperm count of yearling male Indigo Buntings is equal to that of adults, and yearlings are equally likely to find mating opportunities. In the case of Lazuli Buntings, adult males are less aggressive to second-year males than to older birds. Both very bright males and dull-coloured males tend to gain better territories than those of intermediate-plumaged birds (see Breeding), suggesting that delayed plumage maturation confers an advantage on yearlings by allowing them to mate without attracting adult aggression. Further, it has been suggested that delayed plumage maturation allows young males to mimic females on the wintering grounds and thereby reduce competition with adult males. V. R. Muehter and his colleagues, however, showed that adult male Lazuli Buntings could discriminate between second-year males and females, rendering the female-mimicry hypothesis unlikely for that species, at least.

Saltators, which probably do not belong in the family Cardinalidae (see Systematics), are about the size of aCatharus thrush, the Green-winged Saltator weighing 38-46 g. They have a relatively long tail and relatively long, strong legs, as reflected in the genus nameSaltator, which means "dancer" or "jumper". No saltator is brightly coloured, although some have a red or orange bill and/or a striking head and throat pattern, including white superciliaries and a black-bordered throat patch. With the exception of the Slate-colored and Black-throated Grosbeaks and the Thick-billed Saltators, which have limited differences between male and female,Saltator species lack sexual plumage dimorphism. Some species, including the Golden-billed Saltator, have duller juvenile plumages, sometimes with streaked underparts. The bills of saltators vary considerably in length and relative thickness. That of the Thick-billed Saltator is, as its English name implies, particularly short and stout compared with the bills of other members of the genus.

The genus, as represented by the Buff-throated Saltator (Saltator maximus), has a well-developed M. coracobranchialis cranialis, a shoulder muscle that is vestigial in most nine-primaried oscines, but well-developed in the paulid genusZeledonia and the thraupidCatamblyrhynchus. R. J. Raikow considered the condition in these genera to be primitive for the group. In contrast, the dorsal belly of M. obturatorius longus is small, a condition that Raikow considered to be derived; the muscle is large, the presumed basal condition, in the genera Pheucticus, Cardinalis andPasserina.

Although some saltators include leaves in the diet (see Food and Feeding), their gut morphology is similar to that of other frugivorous birds. Their only apparent adaptation to leaf-eating may be a serrated inner edge on the maxilla. Another possible morphological consequence of their largely frugivorous diet may be the fact, noted by A. Smith, that saltator carcasses decay particularly rapidly after the birds have been killed.



Cardinalid habitats range from Amazonian rainforest toAndean forests, semi-arid scrub, temperate woodland and, for the Dickcissel, grassland and open prairie. Most species have broad habitat tolerances, generally preferring semi-open and edge situations with shrub or tree cover. Localized habitat differences can affect survival. As an example, Black-headed Grosbeaks in the Sacramento Valley of California, in the south-west United States, survived less well at a site that had been grazed by cattle and had, as a result, a less dense shrub community than they did at three other localities in the region.

A common feature ofmany cardinalid habitats, those occupied by Dickcissels being a possible exception, is a high degree of complexity in vegetation shape and structure. S. P. Brennan and Schnell found that abundances of Northern Cardinals, Blue Grosbeaks and Indigo and Painted Buntings recorded on Breeding Bird Survey transects in the central plains of the United States were correlated with fractal dimension, a measure of the shape complexity of habitat patches. Northern Cardinals and Painted Buntings reached their greatest density in areas with irregularly shaped patches, while abundance of all except the Indigo Bunting was correlated positively with edge density, a measure of the amount of edge in a habitat segment.

J. H. Vega Rivera and his colleagues analysed the habitat requirements of the Orange-breasted Bunting, a species confined to western Mexico, where less than 1% of itsdry-forest habitat is under any form of protection regime.They found that the most important parameters for predicting the presence of the bunting werehigh seasonality in rainfall, low diurnal variation in temperature of no more than 9·5-15·5°C, and low elevation, from sea-level to 900 m. Within these parameters, the buntings occupied areas of forest varying widely in the extent of cover that they provided.

In the United States,Indigo Buntings are historically inhabitants of successional-stage habitats, and have expanded into forest areas in southern Appalachia as new highways created additional edge situations. In north-east Texas, Indigo and Painted Buntings studied by J. G. Kopachena and C. J. Crist overlapped broadly in habitat choice, but differed in respect of habitat heterogeneity. Both are species of edge habitat, but Indigo Buntings were more likely to be found in openings in otherwise wooded areas, while Painted Buntings preferred clumps of trees in otherwise open habitat. Painted Bunting habitats provided subtle, natural edges, rather than the abrupt edges along forest clearings apparently preferred by Indigo Buntings. There seemed to be no difference between the two species in the vegetational structure of song perches that they used, although the tree species involved did differ.

Apart from the probably non-cardinalid generaParkerthraustes andSaltator (see below), only the twoCaryothraustes species and the singlePeriporyphus grosbeak are primarily forest-dwellers.Caryothraustes grosbeaks are inhabitants of the forest canopy, although the Black-faced Grosbeak will forage at lower levels in more open situations, such as shaded cacao plantations. The locally distributed Red-and-black Grosbeak keeps to lower and middle strata in the forest interior.

The dispersal of cardinalids, includingCaryothraustes andPeriporphyrus, in Amazonian forests has not been impeded by large rivers. In north-west Argentina, Black-backed Grosbeaks can cross habitat gaps between relict montane forest patches, presumably because they can occupy also non-forest habitats in the region. The Blue-black Grosbeak (Cyanocompsa cyanoides) is a bird of dense undergrowth, primarily in forest, but it will forage and nest hundreds of metres beyond the forest edge, even in cultivated maize (Zea mays) fields, so long as there is sufficient cover.

Some species, particularly the Vermilion Cardinal, Pyrrhuloxia and Varied Bunting, and Venezuelan populations of the Ultramarine Grosbeak, live in arid country, while the Black-backed Grosbeak prefers xeric open woodland with underbrush, usually in mountainous areas. Varied Buntings are typically inhabitants of dense thorn-scrub in canyons and arroyos (creeks), and can live in open desert provided there is a dense cover of mesquite (Prosopis), acacia (Acacia) and other woody shrubs. They do not require a source of open water. Typical habitat at the south-western edge of the Edwards Plateau, in Texas, usually includes scattered trees, which provide perches for territorial males. The Vermilion Cardinal is an arid-zone specialist whose abundance in undisturbed sites is correlated with the height of the vegetation, presumably because it moves through different levels while foraging and usually sings from the tops of columnar cacti. Genetic studies by A. Rodrí­guez-Ferraro suggest that it may have expanded into its range in arid regions of northern South America from a population base on Margarita Island, off north Venezuela. Encroachment by velvet mesquite (Prosopis velutina) into grassland habitats in south-eastern Arizona has favoured Pyrrhuloxias, which often use mesquites as nest-sites. Pyrrhuloxias have a slightly greater ability than do Northern Cardinals to reduce loss of body mass during evaporative cooling, and this may give them a physiological advantage in hot, dry areas.

Dickcissels are obligate grassland specialists, but within that category they have specific habitat preferences. They prefer dense, moderately tall vegetation, with deep litter and a high proportion of forbs to provide nesting cover and support. Unlike some other grassland specialists, Dickcissels do not sing in flight, and as a consequence they require many elevated song perches. Optimal habitats include mature old fields, hay fields, fencerows, hedgerows, and moderately grazed to idle prairie.

On their breeding grounds, Dickcissels prefer the moister portions of grassy areas, including meadows and man-made habitats of clover (Trifolium), alfalfa (Medicago sativa) and other low-growing crops. They are more sparsely distributed in arid areas, and may have spread into eastern Colorado only after irrigation made possible the growing of alfalfa. R. S. Mulvihill has suggested that their regular invasions of eastern North America, including the establishment of isolated breeding populations along the eastern seaboard, may be a response to drought in the core portion of the species' range. D. B. McNair, however, failed to detect any correlation between invasions and climate (see Status and Conservation).

In Kansas, Dickcissel abundance is best predicted by examination of the vertical profile of the habitat, which relates to the availability of male song perches, while nest success is best predicted by the density of vegetation and litter cover. J. L. Zimmerman has proposed that the preference of Dickcissels for old fields is not driven by any advantage that this habitat confers on female nesting success. Old-field habitat does, however, provide a greater assortment of suitable nest-sites than a more homogeneous grassland would, and, since the Dickcissel is polygynous, a male can sequester a greater number of females within an old-field territory than he can in a territory on open prairie.

Dickcissel nests are more likely to be preyed on or parasitized if they are within 50 m of tree cover. This may contribute in some areas to the avoidance of woodland-edge habitats, although cropland edge apparently provides no such disadvantage. Dickcissels on reclaimed mining lands in Texas were more likely toselect nest-sites farther from riparian areas, where cowbird (Molothrus) parasitism was at its highest, and closer to areas encroached by brush.In the Flint Hills of Kansas and Oklahoma, they reach their highest density in grasslands restored under the Conservation Reserve Program (CRP), but they have greater nesting success in unburnt hay fields, where the density is about half of that on CRP sites and the rate of cowbird parasitism is lower.

D. E. Burhans found that Indigo Buntings in Missouri were at greater risk from Brown-headed Cowbirds (Molothrus ater) in forest, where female cowbirds prefer to look for suitable hosts, than in old-field habitats. Within old fields, parasitism was lower at better-concealed nest-sites.

For migratory species, wintering habitat is often the same as, or similar to, the habitat that they occupy during the breeding season. Indigo Buntings, however, may winter in grasslands and crop fields, sometimes in association with such other avian species as grassquits (Tiaris). Large flocks, sometimes containing hundreds of individuals, feed in harvested rice fields in Belize. In La Sepultura Biosphere Reserve, in the south Mexican state of Chiapas, where wintering Indigo and Painted Buntings overlap with resident Rose-bellied and Orange-breasted Buntings, the resident species are restricted to tropical semi-deciduous forest, whereas the migrants have a much wider habitat range and also occupy pine (Pinus) and pine-oak forests and cloudforest.

E. Martí­nez-Meyer and his colleagues modelled the ecological niches of Passerina buntings on their breeding and wintering grounds. They concluded that in this genus, which presumably arose in Middle America, where it reaches its highest species diversity, wintering habitats are conservative while breeding-ground habitats are derived. The characteristics of winter habitats should also, therefore, be found in breeding habitats, but not necessarily vice versa. These scientists designated Lazuli and Varied Buntings as conservative "niche-followers" whose breeding habitat resembles that occupied in winter. Indigo Buntings, Painted Buntings and Blue Grosbeaks have undergone a shift in their choice of breeding habitat, so that, although characteristics of the winter habitat are good predictors of their breeding distribution, the reverse is not the case.

Pyrrhuloxias in Texas occupy winter roosts in dense mesquite thickets. Winter flocks of, at times, up to 1000 individuals feed on road shoulders, along fencerows, in weedy fields and at field edges. On their wintering grounds Dickcissels roost in huge numbers in sugar-cane (Saccharum) fields or tall stands of native grass, and feed primarily in cultivated rice and sorghum fields and, to a lesser extent, in native grasslands; they repair to fallow, bushy vegetation near water in order to rest between foraging bouts. All three types of habitat must be present to support a massive wintering flock of this species (see General Habits, and Status and Conservation).

Both in their core range in Tennessee and in newly occupied areas in Ontario, Northern Cardinals occupy a wide range of types of woody cover. Some habitats that in Ontario supported only low population densities of the species in the 1960s held considerably higher concentrations in Tennessee. D. D. Dow suggested that the northward expansion of the species may have been limited by heavy winter snowfall, rather than by vegetation type.

In eastern Texas, Northern Cardinal territories are larger in areas having mature trees than in areas with only shrubby vegetation. This is presumably because the birds have a strong affinity for understorey foliage and because shrub cover is lower where large trees are present. In central Ohio, cardinals are more abundant in urban forests than in rural riparian forests, but are equally successful in both. Urban development, in Ohio at least, has apparently favoured cardinals by creating areas with dense exotic shrub cover, warmer minimum winter temperatures, and ready access to garden bird feeders.

In this case the urban landscapehas not functioned as an ecological "trap", luring the birds into an area where they would fare poorly. The abundance of anthropogenic edge habitat, however, may create such a trap for Indigo Buntings. A. J. Weldon and N. M. Haddad, using experimental forest plots in South Carolina, found that these birds preferred patches with increased areas of edge, but did less well within them. Plots with greater edge areas experienced higher rates of nest predation in the first half of the breeding season; Indigo Buntings breeding within them fledged significantly fewer young than did pairs in rectangular plots having the same area, even though predation increased in the rectangular plots in the late nesting period.

Farther north, in south Canada, tree-harvesting in woodlots in southern Ontario appeared to have no effect on the breeding success of Rose-breasted Grosbeaks, as measured by nest survival, numbers of young fledged and other parameters. Woodlots which had suffered heavy removal of wood had more fruit-bearing trees, and higher densities of grosbeaks, than did undisturbed sites. In all of the woodlots studied, both disturbed and undisturbed, however, the grosbeaks failed to produce enough young to replace themselves. Populations of Rose-breasted Grosbeaks require large tracts of continuous forest in order to thrive. Overharvested woodlots, with their increased food supplies, may also act as ecological traps, attracting birds into a population sink.

Northern Cardinals introduced in Hawaii are found throughout disturbed lowland forests and gardens, and also, less commonly, in native forest. They are most common in dry, open, lowland forests with a grassy or shrubby understorey. They have reached the Alaka'i Swamp, at 1200 m elevation on Kauai, and there is a record of the birds at 4055 m near the summit of Mauna Kea, on the main island of Hawaii. Some populations may nonetheless have declined since the 1970s. Those in open forests in the Hakalau Forest National Wildlife Refuge, on the island of Hawaii, showed a stable to declining trend in the 21 years after the refuge was established, in 1987. In the Puna forests on the same island, cardinals have benefited from the widespread presence of illegal marijuana (Cannabis sativa) plantations, which provide them with canopy openings and sources of seed.

Mention must be made also of the two anomalous, probably non-cardinalid generaParkerthraustes andSaltator. The first of these contains only one species, the Yellow-shouldered Grosbeak, which is an inhabitant of tropical forest interiors. The 16 species ofSaltator range widely in habitat preferences, although most are birds of arid to semi-arid open country, including second growth and farmland. Green-winged Saltators in Argentina have recently spread into Entre Rios and the north-eastern part of Buenos Aires Province, where they are frequently found in farmlands. Incerrado habitat inSão Paulo State, in south-east Brazil, G. Levy found thatBlack-throated Saltators occupied grassland savanna with low shrub density in preference to more heavily vegetated areas. Trees greater than 2 m in height, which the saltators use as sentinel posts, are important elements in the habitat.

Buff-throated Saltatorsin Costa Rica are denizens of second-growth thickets and forest edge, including forest treetops near the borders of clearings. In contrast, the Streaked Saltator avoids areas with continuous tree cover, being apparently restricted to areas of tangled vegetation, such as overgrown pastures. Where Green-winged and Greyish Saltators occur together inchaco vegetation, the former is found mainly in heavy brush bordering rivers, whereas Greyish Saltators occupy dense brushy areas in woodland, forest, savanna and grassland.

Some saltators, includingthe Slate-colored and Black-throated Grosbeaks and the Black-winged Saltator (Saltator atripennis), are primarily forest-dwellers. The rare and localized Masked Saltator occurs in humid montane forests in Ecuador and Peru, where it seems to depend largely on stands ofPodocarpus oleifolius. In Colombia, these stands make up less than 10% of the total forest area within this saltator's range, so that the species occurs at very low densities in that country (see Food and Feeding, and Status and Conservation). The Rufous-bellied Saltator, confined to Bolivia and north-western Argentina, is a high-altitude specialist, found in montane scrub generally above 2500 m, includingPolylepis woodland and adjacent cultivated areas.


General Habits

There is little information on the amount of time that cardinalids budget for daily activities. Captive cardinals spent 74% of their time in perching, 13% in feeding and 5% in flying. In the United States, male Dickcissels on their breeding grounds in Kansas spent 17-21% of their time in foraging, 8-10% in resting, and at least 50% in singing, the last figure rising to more than 70% during the major influx of returning females in spring. Other Dickcissel activities varied in intensity throughout the breeding season; territory defence occupied 3·5% of the birds' time during the second week, but diminished to near zero as the season progressed. The time devoted by females to maintenance activities rose during the incubation and brooding periods. Time budgets were affected by ambient temperature. At temperatures above 94°F (34·4°C) Dickcissels spent more time in resting, especially if humidity was high; below 50-60°F (10-15·6°C) they increased their evening foraging bouts, presumably to gain enough energy to maintain body temperature through the night.

The flight of cardinalids is strong,and for many species it is also undulating, but that of Indigo Buntings and Dickcissels is direct. Cardinalids fly or hop from branch to branch in trees and bushes, and Blue Grosbeaks may sidle rapidly along a branch in the manner of a parrot. When on the ground, cardinalids generally hop, rather than walk, although Dickcissels both hop and walk and Painted Buntings may include walking in some of their displays.

Like other passerines, cardinalids preen regularly. Rose-breasted Grosbeaks, in addition to wiping the bill, ruffle their head feathers and rub the side of the face against a perch, especially after bathing. So far as is known, cardinalids, except very young birds, perform head-scratching by the indirect method, one leg being brought over the wing. When no more than nine days old, young Rose-breasted Grosbeaks may attempt to scratch the head with a leg held over the wing, often losing their balance in the process.

When water is available, bathing is a frequent maintenance activity of the members of this family, Rose-breasted Grosbeaks, for example, bathing at least once a day when possible. In contrast, the Dickcissel is the only member of the family that has been recorded as dust-bathing. Anting has been recorded for the Northern Cardinal, the Black-faced Grosbeak and the Dickcissel. Northern Cardinals, Rose-breasted Grosbeaks and, at times, Indigo and Varied Buntings sun-bathe. Northern Cardinals indulge in this activity even on very warm days, and it has been suggested that the extra heat may make feather parasites more active and easier for the bird to remove while preening.

The roosting habits of several of the cardinalids are known. Northern Cardinals and Pyrrhuloxias roost in dense shrubs or thickets. In addition to nocturnal roosting, Northern Cardinals may spend a substantial portion of the daylight hours at a roost, resting and perhaps occasionally sleeping. Their roosts are not shared. Pyrrhuloxias in Texas settle in roosts in mesquite-hackberry (Prosopis-Celtis) thickets from approximately half an hour after sunset to just before total darkness. Indigo Buntings have been seen to fly into dense Rubus bushes at dusk during the breeding season; on their wintering grounds, they roost in tall grass, in marshes and in trees.

The most expansive roosts in the family, however, are those of Dickcissels on their wintering grounds, where the numbers involved can be enormous. Roosts of between 20,000 and 2,950,000 individuals have been recorded in Venezuela, with an average roost size of 580,000. Roost-sites may be used for several years. Most birds remain in the same roost throughout the winter, although some shifting from roost to roost may occur. Most of this species' roosts are in fields of sugar cane, although the birds may be forced to move when the plantations are harvested or rotated to other crops. They may also roost among cat-tails (Typha), in tall grass or shrubs, in small trees and in fallow rice fields. During the spring migration, thousands of Barn Swallows (Hirundo rustica), Collared Sand Martins (Riparia riparia) and Bobolinks (Dolichonyx oryzivorus) sometimes join Dickcissels at their roosts.

Dickcissels observed in Venezuela arrived at the roost about90 minutes before sunset and began to depart 10-30 minutes after sunrise. In late March, shortly before migration, they left earlier and returned later. Roosting individuals perched on sugar-cane leaves or crowded together on the ground. It took nearly an hour for large roosts to fill up with birds, but in the morning the same roosts would empty in less than 30 minutes as hundreds of thousands of Dickcissels left together, spiralling upwards into tornado-shaped funnels or broad columns.

In addition to their enormous nocturnal roosts, Dickcissels gather close to their feeding grounds in daytime roosts containing from a few tens to hundreds of thousands of individuals. Here, they rest between morning and late-afternoon foraging bouts. Only rarely were daytime roosts sited in sugar cane, the preferred habitat for nocturnal roosts.

Roosts of this species in Trinidad, though considerably smaller than those in Venezuela, are still substantial. R. ffrench conservatively estimated that one roost contained some 66,000 individuals, and on another occasion he saw perhaps 100,000 birds fly out of the cane field within a period of little more than two minutes. The Dickcissels usually flew to their roosts at a height of about 60-90 m above the ground, breaking formation above the roost and cascading into it with a great deal of noise as the wind rushed through their flight-feathers. As in Venezuela, they normally left their roosts about 25 minutes after sunrise, their departure time becoming earlier in the three weeks before the spring migration. Unlike those observed in Venezuela, however, they did not perch on the ground while roosting but, instead, stayed at least 2 m up in the sugar cane, perhaps in order to avoid snakes, rats and introduced mongooses (Herpestidae).

True social behaviour is uncommon in the Cardinalidae. Black-faced and Yellow-green Grosbeaks are regularly found throughout the year in groups of up to a dozen or more, sometimes joining with other species. They apparently do not defend territories. The groups may consist of members of an extended family, larger groups being assemblages of family units (see Breeding).

Gaping is an aggressive displayused by both Northern Cardinals and Rose-breasted Grosbeaks. In the case of the cardinals, it forms part of the "head-forward display" used at higher levels of aggression, in which an individual faces his opponent, lowers his body, flattens his crest, opens his bill, and may quiver or spread his wings. While giving this display, the bird may be silent or may give a variety of "chip" notes and other calls (see Voice). Rose-breasted Grosbeaks perform a similar display.

Alarmed or arousedLazuli, Indigo and Varied Buntings perch upright, with the body feathers sleeked and the crown feathers raised. Among Indigo Buntings, the most intense response to playback of the species' recorded song appears to be a "moth flight" display, in which the male quickly flies towards the source of the sound, fluttering his extended wings as he lands on a nearby perch, and crouches with fluffed body plumage, lowered and rapidly quivering wings and raised beak and tail.

S. M. Lanyon and C. F. Thomson described a repertoire of displays by Painted Buntings, most of which are performed also, with variations, by Indigo and Lazuli Buntings. They include, among other things, upright posture, bowing (a display apparently lacking in Indigo and Lazuli Buntings), wing-quivering, and fluffing of the body feathers, and are used in aggressive encounters, as well as in courtship (see Breeding).

Among ten North American cardinalid species, the maximum recorded longevity in the wild ranges from 4 years, for the Dickcissel, to 15 years 9 months, for the Northern Cardinal, with a mean maximum of 8-9 years. Two Indigo Buntings returned to A. Downer's feeding station in Jamaica for eight successive years. These figures probably do not, however, reflect either average life expectancy or potential longevity. R. B. Payne and L. L. Payne estimated the mean survival rate of colour-ringed Indigo Buntings in southern Michigan to be only 2·04-2·41 years; the mean survival of individuals recovered from ringing programmes fell within these limits. At the other extreme, the oldest Rose-breasted Grosbeak recovered in the wild was 12 years 11 months in age, but captive specimens have lived to an age of 24 years. Similarly, the maximum recorded longevity in the wild for the Black-headed Grosbeak is 9 years 1 month, but a captive specimen lived for 25 years. In the case of Pyrrhuloxias, males may have higher survival rates than females: in an unpublished study, of 44 individuals recaptured more than a year after ringing, 73% were males.

Domestic cats and hawks (Accipitridae) and owls (Strigidae) of various species have been recorded as taking cardinalids as prey. Pyrrhuloxias form a small part of the diet of Ferruginous Pygmy-owls (Glaucidium brasilianum) in Texas, and Great Grey Shrikes (Lanius excubitor) attack and kill Northern Cardinals. Rose-breasted Grosbeaks may be seized by Sharp-shinned (Accipiter striatus) and Cooper's Hawks (Accipiter cooperii). A grey squirrel (Sciurus carolinensis) has been recorded as killing and eating an adult Northern Cardinal. Northern Cardinals adjust their foraging strategy to the potential presence of a predator; in one experiment, cardinals that were offered high-quality food at feeders away from cover visited these more frequently than they visited feeders with lower-quality food near cover, but they reversed this preference if exposed to calls of a predator, Cooper's Hawk, at two-hour intervals.

The enormous winter roosts of Dickcissels are magnets for predators. In Trinidad, ffrench noted Merlins (Falco columbarius), Common Barn-owls (Tyto alba) and an Aplomado Falcon (Falco femoralis) hunting at Dickcissel roosts. Further, at roosts in Venezuela, G. D. Basili and S. A. Temple recorded five species of falcon, seven hawks, two owls and three mammals as preying on Dickcissels; the mammals were the jaguarundi (Puma yagouaroundi), greater grison (Galictis vittata) and tayra (Eira barbara). Merlins were the most common diurnal predators, with 1-5 birds at each roost, and Common Barn-owls the most conspicuous by night.

Saltators, probably not true cardinalids (see Systematics), are in general not particularly well known. In thellanos of Venezuela, Greyish and Orinocan Saltators were found to spend 77% to 80% of their time in sitting, singing and preening. Buff-throated and Green-winged Saltators have been recorded as using anting behaviour, and A. F. Skutch observed a Buff-throated Saltator as it visited an arboreal ant nest for this purpose.

It appears that the social systems and display repertoires of saltators have never been thoroughly studied. Black-headed (Saltator atriceps) and Buff-throated Saltators are regularly found in small, loose groups, but whether the members of these groups are related to one another is not known. Other species, including the Greyish and Streaked Saltators, are usually found only singly or in twos and threes. H. Sick described a display of the Black-throated Grosbeak in which the bird opens its wings to show off the white underwing-coverts, but he did not mention the context in which this display occurs.

Black-throated Saltators associate in groups of 2-6 individuals. Helpers at the nest have been recorded for this species, suggesting that these groups consist of family-members. One individual, very occasionally accompanied by one other, usually sits sentinel on a prominent exposed perch while other members of the group forage near the ground. Sentinels studied byJ. Ragusa-Nettoin the Braziliancerrado stayed at their posts for usually less than six minutes, but sometimes for as long as 30 minutes, constantly turning the head from one side to the other. Individuals spent longer at their post if there were a large number of raptors in the area. If a bird of prey came into view, the sentinel sounded an alarm call, causing all fellow group-members to head for cover. Except during the breeding season, Black-throated Saltators constitute one of four nuclear species in mixed flocks in thecerrado. Up to 33 other species may be present in a saltator-centred flock, these other species taking advantage of the warnings that saltator sentinels provide. While Ragusa-Netto recorded attacks on mixed flocks by American Kestrels (Falco sparverius) and Aplomado Falcons, he never saw an attack on a group consisting of saltators alone.

Little is known about predation suffered by saltators. A Golden-billed Saltator was among the prey items found in the nest of a Ferruginous Pygmy-owl in Argentina, but no other relevant information is available.



The voices of cardinalids are among their most attractive features. Most species have musical songs consisting of a series of whistled or warbled phrases, as well as a repertoire of often sharp and distinctive calls. The primary song of the Dickcissel is an exception, being a short, dry series of notes from which the bird's English name is derived.

Very little research has been undertaken on the vocalizations of Mexican, Central American and South American cardinalids, some of which appear to have complex song repertoires. For example, Skutch describes the Blue-black Grosbeak as "a superb and generous songster". This species, the song of which varies geographically across its wide range, is deserving of further study. The singingbehaviour of some North American species has, however, been the subject of intensive research, particularly by R. E. Lemon and G. Ritchison on Northern Cardinals, as well as recent work on the acoustics and neurophysiology of cardinal song by N. Fletcher and his colleagues, and studies by Ritchison on Black-headed Grosbeaks, by K. D. Groschupf and C. W. Thompson on Varied Buntings, and by R. B. Payne and his students at the University of Michigan on Indigo Buntings.

O. N. Larsen and F. Goller, in their investigation of song production in the syrinx of the Northern Cardinal, used endoscopic examination of the syrinx and electrical stimulation of syringeal muscles in living birds. Their studies represented the first direct examinations of the oscine syrinx in action. They demonstrated that, contrary to earlier belief, the medial tympaniform membranes are not the principal sound-generators in the syrinx; birds were able to produce nearly normal song even if these membranes were incapacitated. Instead, sound is generated primarily by the syringeal labia. The syrinx is controlled, according to Larsen and Goller, by complex interaction between muscles in combination with the complicated suspension of the cartilaginous framework upon which they act. Larsen and Goller found no significant difference in the operation of the syrinx between the Northern Cardinal and the American Crow (Corvus brachyrhynchos) and Brown Thrasher (Toxostoma rufum), the only other oscines that they examined. It is likely, therefore, that the syrinx in cardinalids functions in a manner typical for songbirds in general. The left side of the syrinx produces lower frequencies, up to about 3·5 kHz. Both sides are used for higher frequencies, but airflow is often reduced or even eliminated on the left side, suggesting that the right side specializes in producing higher-frequency sounds. Most song syllables, however, span the frequency ranges of the two sides, and the shift of airflow from one side to the other usually cannot be detected.

Beyond the syrinx, the volume and tonal quality of cardinal songare influenced by the degree of beak gape, and by complex movements of the hyoid apparatus that affect the position of the tongue, larynx and trachea. Northern Cardinals, like other songbirds, possess an oropharyngeal oesophageal cavity between the top of the trachea and the open beak which can be actively distended by movements of the hyoid skeleton. T. Riede and his colleagues have shown that the cavity acts as a resonator, the sounds filtered within it being radiated through the beak. By adjusting its size, the bird can effectively "set" the cavity so that its resonance frequency matches the fundamental tone produced in the syrinx, resulting in a nearly pure-tone song.

Female Northern Cardinals, as well as males, sing, both in duet with their mates and on their own. Jawor and S. A. MacDougall-Shackleton have demonstrated that the HVC, RA, and Area X nuclei in the brains of Northern Cardinals, which form part of the song-control system, vary in volume both seasonally and between the sexes. They are largest early in the breeding season, and male nuclei are 1·5-2 times larger than those in females. Both differences are small compared with those in bird species that have more restricted singing periods or whose females do not sing. In addition, females of a number of other cardinalids are known to sing, as illustrated by the Rose-breasted and Black-headed Grosbeaks and the Black-faced, Red-and-black and Blue-black Grosbeaks; female Pyrrhuloxias sing, albeit rarely, and females of two of the saltator species also do so (see below). Blue-black Grosbeak partners sing to each other while building the nest and while the female is incubating. In contrast, females of the Dickcissel and ofPasserina species are not known to sing.

Northern Cardinals may begin to produce soft warbling soundsat as early as three weeks of age, but true song development does not really get underway until the following winter and spring. Initial warbling is succeeded by disorganized, rambling subsong, which slowly develops into a more controlled plastic song as syllable morphology develops. Plastic song crystallizes into recognizable adult song when the bird reaches about ten months of age.

Some simplecardinalid sounds may be innate, but complex song elements are apparently learnt during a sensitive period early in life. The maintaining of song structure, once learnt, requires memory reinforced by auditory feedback. Experiments demonstrated that the songs of artificially deafened Northern Cardinals gradually deteriorated with time, and deafened juvenile Black-headed Grosbeaks were unable to develop normal song.

Hand-reared Northern Cardinals raised in isolation developed simple, but not more complex, two-note or three-note phrases resembling those of adults, but isolated Indigo Buntings were unable to produce more than a few adult-type syllables. In the case of the Northern Cardinal, both sexes learn song from their parents during their first summer. Females, but not males, require exposure to adult songs early in life in order to learn. Captive male cardinals younger than 15 weeks in age, if exposed to recordings of adults, developed in the following spring songs that copied some or all of the recorded songs; they also invented material of their own. Male Northern Cardinals were also able to learn songs played to them when they were between 5·5 and nine months of age.

A. Yamaguchi, working with captive Northern Cardinals, found that the sensitive period begins at roughly the same time for both sexes, on average 27·8 days for males and 22·4 days for females. The learning period, however, lasts much longer for males, ending at an average age of 214·9 days, the corresponding figure for females being 71·1 days. Despite this, the number of songs in the birds' eventual repertoires was roughly the same. Each sex appeared capable of learning the song of the other, and showed no preference for learning the song of its own sex. The difference between the sexes may be related to the need for the male, but not the female, to be able to learn the songs of neighbouring conspecifics after it settles on territory.

Male Indigo Buntings can learn syllables from recorded songs up to 18 months of age. In Indigo and Lazuli Buntings, the patterns of song-learning are quite different. Female buntings do not sing, and males learn their songs not from their parents, but from other males, although it is possible that they may learn some basic elements of song from hearing their fathers. Fledgling Indigo Buntings between 15 and 60 days old begin to sing highly variable, apparently random subsong; this crystallizes into adult song during the first spring, about 9-11 months after the bird hatched. Lazuli Buntings produce subsong at 29 days, but stop singing in winter and begin subsong again in early spring.

Approximately 80% of male Indigo Buntings copy the song of another male, usually an adult, rather than another yearling, that they encounter when establishing their territories during their first spring season. As the birds normally settle several territories away from their natal area, only very rarely is the song that of their father, or, indeed, that of any close relative. Members of a song neighbourhood, defined as an area in which all males sing essentially the same song, are, by and large, unrelated males.

The advantages of learning froma neighbouring bird may be considerable. For example, first-year individuals that copy a neighbour's song have greater success in mating, nesting and bringing off fledged young than those that do not. Possibly, they are able to deceive older birds, which may otherwise chase them off, into thinking that they are established territory-holders. Adults whose songs match those of their neighbours also tend to be more successful than adults that are not part of a song neighbourhood. The breeding success of Indigo Bunting males, however, seems not to depend on the specific song neighbourhood to which a bird belongs.

The choice of whichindividual bird to copy, out of a number of singing males that may be present, apparently requires intensive social interaction between learner and tutor. These interactions can include repeated chases across territorial boundaries and also countersinging, the younger bird matching the song of the older one.

Apparently, the 20% of Indigo Buntings that do not copy from a neighbour learn their first song before arriving on their territory. Some first-year males, and approximately 5% of older males, may switch songs in the course of a season, either when changing territories or, for older birds, if no neighbouring males match their current song. Adult Indigo Buntings appear to retain the ability to learn a new song even after their original song crystallizes, permitting them to replace it with another that matches that of a neighbouring male. This relearning ability contributes to the development and differentiation of song neighbourhoods.

Lazuli Buntingsin their first spring may copy song segments from as many as ten other males before settling on a stable song. Second-year Painted Buntings rarely establish territories, but they may sing from within the territory of another male; in this instance, the pattern of song-learning is not known.

In general, the song repertoires of cardinalids contain a number of syllables, sometimes composed of more than one subunit, arranged in various combinations to produce different song types. The number of syllables used in any one song varies among species. Pyrrhuloxias frequently use only a single syllable, while Rose-breasted Grosbeaks may incorporate 9-11 syllables into a single song.

The song of the Blue-black Grosbeak has not been studied in detail, but Skutchand others have noted that it is frequently in two parts, a series of often descending tonal whistles being followed by a rapid warble at lower volume. The concluding warble is not audible at so great a distance as is the first part of the song, suggesting that it may function only in close encounters.

Syllable construction, length and amplitude vary among members of the family. In studies in Canada, most of the 14 syllables identified in the songs of Northern Cardinals in London, in south Ontario, were simple whistles, frequently slurred downwards or upwards, the lower frequencies ranging between 1·2 kHz and 1·6 kHz, but the upper frequencies varying more widely, from 1·2 kHz to 6·2 kHz. Mean syllable length varied from 0·10 seconds to 0·48 seconds. Rose-breasted Grosbeak syllables in Quebec ranged in frequency from 1·5 kHz to 4·0 kHz and had an average duration inthe range 0·24-0·28 seconds.

When incorporated into a song, the intervals between successive syllables given by Northern Cardinals and Rose-breasted Grosbeaks relate directly to the duration of the preceding syllable. With Rose-breasted Grosbeaks, the interval is some three times longer than the syllable itself. Syllables in the songs of Painted and Orange-breasted Buntings are evenly spaced, whereas the intervals in songs of Indigo, Lazuli and Varied Buntings are more irregular. Syllables may be repeated in sequence.Cardinalis may repeat the same syllable a number of times, and Indigo Buntings normally sing each syllable at least twice before moving to the next. Rose-breasted Grosbeaks rarely repeat syllables consecutively.

The number of syllables in a song repertoire varies.Cardinalis may have only ten or so syllables available; in various studies, 8-21 were recorded for male Northern Cardinals, 8-10 being a common total, and 1-14 have been recorded for male Pyrrhuloxias. Pheucticus species, so far as is known, generally have larger repertoires: 15-23 for Rose-breasted Grosbeaks and 14-37 for male Black-headed Grosbeaks.Passerina buntings studied have much larger repertoires, of over 100 syllables: 127 are recorded for Indigo Buntings, more than 140 for Lazuli Buntings and 101 for Varied Buntings. In contrast, the song of the Dickcissel consists of repetitions of only two phrases, "dick" and "cissel", occasionally with a third, a trill. Most male Dickcissels sing only a single song type, with slight variations in the number of elements. They repeat the same song throughout the breeding season, and return to it in successive years.

Northern Cardinalsorganize their syllable repertoires into 8-12 song types. These can vary significantly in terms of the number of notes per song. The songs of the conspecific Pyrrhuloxia are organized in a similar manner, although Pyrrhuloxia song syllables are generally shorter. Pyrrhuloxias have a higher number of song types than do cardinals, but they are mostly repetitions of a single syllable, rather than, as is common among Northern Cardinals, consisting of two or three syllables per song.

In the songs of both cardinals and Rose-breasted Grosbeaks, the order of syllables is not random. Short syllables tend to be repeated more often than longer ones during a single song. Certain syllables are used most frequently at the beginning of a song, and cardinals have distinctive syllables, including a dry "chirr", normally heard only last in the sequence. The "chirr" note may be a measure of fitness. Cardinals are able to replenish their air supply by taking extremely short "mini breaths" if syllables are repeated at rates lower than 16 per second. The elements of the "chirr" follow each other more rapidly than that, so that a male's ability to produce and prolong the "chirr" may be an indicator of his lung capacity and cardiovascular fitness.

Individual repertoires, once established, may persist unchanged for long periods, both through the breeding season and from year to year. Male Northern Cardinals have been recorded as singing the same song types for four years.

Northern Cardinals may repeat the same song for a number of minutes, often with slight variations in volume and structure, but normally beginning with the same syllable. Such a sequence is referred to as a song "bout". Songs within a bout are separated by intervals of varying lengths. Cardinal singing bouts in Kentucky contained an average of 18·5 songs each. When an individual does switch to a different song type, its choice may be determined to a great extent by the song that has preceded it, so that the structure of singing over long periods of time is highly organized. Males in Kentucky varied in the number of notes per song and the number of songs per bout, perhaps reflecting differences in male quality. Oddly, in eastern Texas, Northern Cardinals having smaller repertoires and less complex song types held higher-quality territories and fledged more young, the reverse of what one might expect.

Male and female Northern Cardinals have similar song-syllable repertoires and combine them in the same way. Yamaguchi has shown that male and female songs are not identical, and that the birds themselves can distinguish the sex of the singer. Females may sing longer songs on the nest. Male cardinals repeat syllables in their repertoire in a more stereotyped manner than do females, and female songs have a greater harmonic amplitude than do those of males.

Female Black-headed Grosbeaks have smaller song-syllable repertoires, on average 12·8 syllables, as opposed to 24·5 for males, and their songs, with a mean of 4·5 syllables per song, are simpler and shorter than those of males, which have a mean of 10·3 syllables per song. Females tend to use fewer syllables from their available repertoire in each song. In one example, a female had a repertoire of 14 syllables, but just three of these accounted for 82% of the syllables that she used in her song.

Testosterone may functionin controlling the precision and stereotypy of male songs. The songs of female Black-headed Grosbeaks are more variable than those of males, perhaps as a result of reduced testosterone levels. Occasionally, a female will sing a male-type song, with shorter syllables separated by smaller intervals, possibly in order to mimic a male (see Breeding). In general, males and females in a population share only about 10% of their song repertoires, but mated partners may share 25% of their syllables.

By selecting among their large repertoires of syllables and rearranging the orderof these within a song sequence,Passerina species can create an enormous variety of individual songs. Unlike Cardinalis andPheucticus, Indigo Buntings appear to lack any stereotyped system for organizing the sequence of syllables within a song, except for a single short note given to open a sequence. Some Lazuli Bunting syllables can occur anywhere in a song, but others tend to occur in specific positions; this species' songs usually end with "buzzes", which are rapidly frequency-modulated elements, or with "beeps", which are usually paired chevron-shaped figures with one overtone.

Indigo and Lazuli Buntings regularly use selected syllables more than once, but Painted and Varied Buntings rarely do so. Only about 1% of syllables are repeated in Painted Bunting songs, and Orange-breasted Buntings apparently never use the same syllable twice within a single song. Although the variety of individual songs that they could produce is potentially extremely large, most Indigo and Lazuli Buntings learn only a single, complex advertising song, the "crystallized song", and retain it throughout life; a few males, however, learn songs from two neighbours and may remain "bilingual". In addition, Indigo Buntings occasionally sing longer songs incorporating extra syllable types, as well as low-amplitude squeaky songs incorporating high-frequency elements, employing the latter in territorial encounters. D. Margoliash and his colleagues have established that adult Indigo Buntings commonly sing squeaky, highly variable "plastic" songs resembling those given by young birds during the critical learning period. These low-amplitude songs tend to be uttered in separate singing bouts later in the day than advertising songs, which are usually given at first light. There is evidence that Indigo Buntings may use these plastic songs to practise and "store" additional, newly learnt song syllables before incorporating them into stereotyped song.

Unlike Indigo and Lazuli Buntings,Painted and Varied Buntings do not develop a single crystallized song. Painted Buntings have 4-5 song types, differing primarily in the alteration of figures in the middle of a song. Individual Varied Buntings have repertoires of some 9-35 syllables, first-year males having smaller repertoires than older birds. From these syllables the birds construct songs of varying length. They may add or remove syllables, particularly near the end of a song. Varied Buntings almost never repeat individual songs in the course of a song bout. This extreme variety makes it difficult to classify song types or to determine differences among individuals. Because birds from different areas differ in the ordering of song syllables, no song is shared among populations.

Vocal mimicry is rare among cardinalids, and very few examples have been documented. An adult Rose-breasted Grosbeak heard while singing at Savoy, in Massachusetts, intermixed, among phrases of its own primary song, imitations of the primary songs and calls of Red-eyed Vireos (Vireo olivaceus), Eastern Towhees (Pipilo erythrophthalmus) and Northern Cardinals, as well as the "mew" call and a scold note of the Grey Catbird (Dumetella carolinensis).

With birds, song structure generally correlates with the acoustic properties of the habitat and, in particular, with the ability of song signals to resist degradation with distance. The relatively unmusical song of the Dickcissel may be a response to the increased carrying properties of harmonic, as opposed to tonal, sounds in open habitats. Tonal sounds carry better in closed habitats, and the song of the Blue-black Grosbeak, an interior-forest bird, is composed of pure tone-like sounds with a narrow frequency range. The forest-dwelling subspeciessterea of the Ultramarine Grosbeak has louder and more complex songs than those of the subspeciesargentina, which lives in semi-open country. The difference in this case may be a compensatory response bysterea to low ambient light in its habitat, where plumage signals may be harder to detect.

A study in eastern Texas by M. E. Anderson and R. N. Connor revealed significant differences in frequency and amplitude, correlated with habitat type, among song syllables of Northern Cardinal populations located only 4·5 km from one another. Individuals in younger tree stands lacking a closed canopy produced sounds with a greater degree of frequency modulation than those voiced by cardinals in older stands. In contrast, a study of Indigo Buntings by R. Hylton and R. D. Godard found no significant difference between the songs of birds in open habitats and those in forest habitats. This may reflect insufficient acoustic differences between the two types of habitat. It may, however, be a result of the social learning system for Indigo Bunting song, in which an individual may copy the song of a neighbour even if that neighbour is in a different habitat type.

P. L. Tubaro and D. A. Lijtmaer found similar patterns in the songs of the Red-and-black Grosbeak,Pheucticus grosbeaks and twelve species ofSaltator (see below). Vocalizations used in short-distance communication, such as the wide-band duet songs of the Greyish Saltator, were an exception, but signal degradation is less of a problem if the recipient is close to the singer. Long-distance songs in open country generally contained more elements, with wider bandwidths, higher frequencies and shorter lengths, than did those from birds in closed habitats. Songs in mixed habitats were intermediate in structure. The songs of birds of the forest interior, where reverberation can cause overlap between the echoes of a note and the following song elements, tended to be composed of whistled notes with slowly modulated frequencies. Tubaro and Lijtmaer suggest that the long, slowly modulated terminal notes in songs of the Black-winged Saltator, Black-throated Grosbeak, Slate-colored Grosbeak, Green-winged Saltator and Red-and-black Grosbeak may benefit from reverberation, which could help to create a louder and longer signal.

Blue-black Grosbeaks may sing throughout the year in lowland forests of southern Central America, but at El General,in Costa Rica, which is more than 600 m above sea level, singing begins only in February or March, reaching its peak during the breeding season, from April to September, and diminishing during the heavy rains of October and November. Song production is subject to the control of hormones, particularly testosterone and luteinizing hormone. Increased singing by laboratory-reared male Northern Cardinals in their first spring was associated with a significant increase in levels of these hormones. The increase was much smaller in females. Wild male cardinals have also been found to have far higher breeding-season levels of testosterone and luteinizing hormone than wild females.

Indigo Buntings begin singing just before or during spring migration, and stop prior to moult and the autumn migration. The singing rate of Painted Buntings declines by more than half between the males' arrival on the breeding grounds and the arrival of the females, and declines by half again after pairing. Male Black-headed Grosbeaks reach their peak of singing between the time of arrival on the breeding grounds and the stage at which egg-laying is completed. The frequency of song drops off sharply from the time of laying until the young fledge, although it may rise to pre-laying levels if the nesting attempt fails. Females of this grosbeak begin to sing only during incubation, and their level of singing increases sharply once the young leave the nest. When the young reach independence, both males and females stop singing.

Dickcissels andNorthern Cardinals may sing throughout the year, but do so more frequently during the breeding season. In Arizona, both cardinals and Pyrrhuloxias begin singing in mid-February, reaching a peak in mid-March, and diminishing by September to a few scattered and incomplete songs. Pyrrhuloxias reduce their singing in the latter stages of incubation, but increase it again after the first brood becomes independent. On the wintering grounds Rose-breasted and Black-headed Grosbeaks rarely sing, andPasserina buntings are not known to do so.

Time of year may affect not only the frequency but also the structure of cardinal song. In the case of the Northern Cardinal, Ritchison found that the number of syllables per song varied through the season, being significantly higher in February, March, April and again in August. The number of songs per bout was also greater early in the season. Female Northern Cardinals sing less frequently than males, and sing at their peak for a shorter period. Like males, they can and do sing occasionally in any month of the year. In a Kentucky population studied by Ritchison, males began singing in January and continued until late August or early September, while females did not sing until March, reached a peak in the weeks just prior to nesting, and were rarely heard to sing after mid-May.

Northern Cardinals begin singing earlier in the day during the breeding season. In the northern United States, A. Leopold and A. E. Enyon recorded cardinals in Wisconsin as singing more than an hour before sunrise in late June, particularly when the skies were clear. Rose-breasted Grosbeaks may start to sing as early as 03:00 hours. Both species sing less frequently, if at all, during the midday hours. Indigo Buntings sing from dawn to dusk, reaching a peak in the early-morning hours before sunrise. Lazuli Buntings sing from about one hour before sunrise until 45 minutes after sunset, with a peak in the morning and another at dusk. Varied Buntings sing at any time of day during the breeding season, with no "dawn chorus" or peak singing period. Similarly, at the beginning of the breeding season, when song activity is most intense, male Dickcissels sing throughout the day, often devoting 50% or more of their time to singing.

Many cardinalids sing from a series of elevated perches, either centrally, near the nest tree, in the territory or along the perimeter of it. Ultramarine and Crimson-collared Grosbeaks reportedly sing from deep cover in a tree or bush, although an extralimital Crimson-collared Grosbeak in Texas sang from exposed dead limbs and treetops.Painted Buntings prefer to sing from cover until the females arrive, but after that they do more than half of their singing from exposed perches.Dickcissels sing persistently from elevated perches within their territory, repeating their fairly simple songs every 3-8 seconds.

Northern Cardinals use specific song perchesmore than 10 m above ground, each territory usually having 4-7 such songposts. They will sing also, usually at lower amplitude, from low shrubbery or the ground. Pyrrhuloxias deliver their songs from a number of perches either in the centre or at the edge of their territories. They use the same perches throughout the breeding season, visiting them in a specific order. Varied Buntings do most of their singing from 3-4 exposed perches at the edge of the territory, but they sing from shade if the temperature rises above 37°C. Male Varied Buntings feeding young sing from a standard perch near the nest before delivering food to the nestlings.

A number of species, includingthe Rose-breasted and Black-headed Grosbeaks, sing from the nest while incubating or brooding. Furthermore, these two grosbeaks will sing during flight displays, and Indigo Buntings may sing during aerial chases with other males.

Among cardinalids, the male's advertising song, as is common in oscines, seems to function in the establishing and maintaining of territory, particularly early in the breeding season. The fine structure of the songs of Indigo Buntings apparently contains a number of cues, including redundant information, that signals the territorial intent of the singer. The placement, spacing and structure of the initial figures in the song may serve to attract attention, while precise details of frequency and spacing may contain specific information enabling other birds to recognize the song as that of a conspecific on territory.

Countersinging and song-matching (replying to a song by giving a similar song) among males is documented for a number of species in the family. Northern Cardinals countersing with neighbouring males, as well as with their mates. As they do so, the number of songs per bout decreases, and the number of syllables per song and the number of syllable types per song increase. Male Pyrrhuloxias in adjacent territories sing either in unison or alternately. Lazuli Buntings sing most vigorously during territorial disputes, and they countersing with neighbouring males throughout the breeding season. Several Varied Bunting males within a canyon may sing at the same time. At times, it appears that one male leads the singing and neighbouring males follow.

Male songmay reinforce the pair-bond. The male Lazuli Bunting, for example, sings quietly close to his mate just before egg-laying. With Black-headed Grosbeaks, the territorial function of male song may give way to pair-bond and family maintenance as the breeding season progresses. After egg-laying is complete, males sing mostly in the presence of their mates and are less aggressive towards other males. Once the young have left the nest, the parent Black-headed Grosbeaks sing near their fledglings.

Female song seems to function primarilyin reinforcing the pair-bond, rather than in defending the territory, although it may serve also to advertise to other females that a male is spoken for. In the case of the Black-headed Grosbeak, female song appears to have no territorial role, as neither males nor females respond to playback of the songs of other females of the species, but it is probably important in reinforcing the pair-bond and may serve to maintain the integrity of the family group.

For some species, however, female song may have a territorial role. For example, Skutch heard a female Blue-black Grosbeak sing during a chase involving two or three other birds. Female cardinals may sing in aggressive encounters with other females, and female Pyrrhuloxias, which sing only rarely, may do so while defending the nest. J. Vondrasek found that female, but not male, Northern Cardinals in newly formed pairs sing more often than do those in pairs that had previously bred. This may be because new partners establishing a territory are more likely to find themselves in aggressive encounters with others of the species.

With both Black-headed Grosbeaks and Northern Cardinals, females normally sing only when their mates are present. Unlike those of some other species, female Northern Cardinals apparently do not sing to re-establish contact with a mate or to advertise for a new one; females were found not to increase their song rates if the males were experimentally removed. Male Northern Cardinals frequently duet with their mates, joining the female after she starts to sing. Duetting may provide an opportunity for members of a pair to learn each other's repertoires.

Female Rose-breasted Grosbeaks sing while nest-building, incubating, brooding,and relieving their mate at the nest, while female Black-headed Grosbeaks sing from the nest in the presence of the mate, in the presence of the mate away from the nest, and from foliage as they seek food for the fledglings. Skutch observed a Blue-black Grosbeak female that sang while nest-building with her mate, and he noted that incubating females sang in response to the mate while sitting. A study by S. L. Halkin suggests that singing from the nest by the female Northern Cardinal may signal to the male that food is, or is not, needed, so that the brightly coloured male can avoid unnecessary visits that could attract the attention of predators. Males approached the nest most frequently when the female sang, but, if she sang a song matching his own, a male was more likely to stay away.

In addition to normal female song, Ritchison twice hearda male-type song given by female Black-headed Grobeaks, on both occasions when the male was long overdue at the nest. Ritchison has speculated that the females may have been trying to lure their mate to the nest by deceiving him into thinking that an intruding male was in the vicinity. He did not record whether the attempts had been successful.

In a study by W. L. Thompson, Painted and Indigo Buntings reacted strongly to songs of their own species, but very little or not at all to the songs of otherPasserina. Spacing and frequency range of song elements appeared to be the most important clues in species recognition. Emlen used artificially modified songs to identify the critical features for species identification in the songs of Indigo Buntings. These proved to be the morphology of individual syllables and the duration of the intervals between them. These are among the least variable aspects of song among individuals, and are therefore presumably of greater value in species recognition, as opposed to individual or neighbourhood recognition.

Indigo and Lazuli Buntings outside the zone ofrange overlap (see Systematics) ignore each other's songs, but within it song distinctiveness of the two species tends to break down. In the zone of overlap, the two species defend territories against each other and each responds aggressively to the song of the other species. M. C. Baker found evidence that syllable structure, rather than timing, is the key feature used by non-sympatric populations to identify their own species' song. This is despite the fact that intervals between dissimilar songs are longer in Indigo Buntings than in Lazuli Buntings. This finding conflicts to some extent with Emlen's results.

Indigo Buntings seem not to discriminate between the songs of individual birds, although a male will be more likely to countersing with another male that sings similar song themes to his own. Under some circumstances, a male may react less aggressively to songs of neighbouring birds than to those of strangers. Male Black-headed Grosbeaks, on the other hand, can distinguish songs of neighbouring males from those of strangers, and, presumably because a stranger is more likely to be a challenger than a neighbour on his own territory, will respond more strongly to it.

Because Indigo Buntings learn their songs from adjacent territorial males, the songs ofindividuals in neighbouring territories may be nearly identical. A song neighbourhood can have as many as 22 males sharing a song theme. Although the same song elements can be found throughout the birds' breeding range, specific sequences of three or more elements are more localized. No such sequences were shared between populations 250 km apart.

Similarly, Varied Buntings develop song neighbourhoods, areas in which neighbouring males share either entire songs or shorter sequences of syllables. Painted Buntings were found to share song "figures". Of 129 figures identified in the songs of 93 individuals of this species from Georgia and Florida, only seven were unique to single birds. Painted Buntings farther west in the United States, at the Welder Wildlife Refuge, in Texas, shared song elements, but not whole songs.

D. M. Schook and his colleagues, who examined song-sharing among Dickcissels, found that males in adjacent territories sang similar songs. Song similarity decreased with distance, especially for the somewhat more complex "cissel" element of the song.

Songs of singing male Northern Cardinals tend to resemble, in terms of syllable structure, syllable choice, use of variant syllables, and the organization of syllables into song types, those of other males in the same locality. The territorial fidelity of Northern Cardinals may contribute to the development and persistence of local dialects. The number of syllables shared among cardinal populations diminishes with distance, although some syllables can be heard throughout North America. Of 27 syllables identified in songs in Ontario, nine were shared with cardinals from Ohio but only four with birds from Mexico and British Honduras. Cardinals are more likely to respond to their own dialect than to songs from more distant regions. They may not respond at all, at least on first hearing, to songs from geographically distant populations.

In the Canal Zone of Panama,Blue-black Grosbeaks living along wide stretches of the Chagres River near Gamboa cannot hear conspecifics on the other side, and the species sings different dialects on each bank. On nearby Barro Colorado Island, a distinctive Blue-black Grosbeak dialect has persisted unchanged since at least 1964.

Some Indigo Bunting song traditions can persist for twenty years or more, across many generations ofthe species. The larger the number of birds in a song neighbourhood, the more likely it is to persist. This species' song themes do, however, change in the course of time. Local immigration and the deaths of males singing certain song types are partly responsible. Some first-year males modify the song that they learn from a neighbour, by adding, subtracting or modifying a song element. These modifications are, in subsequent years, learnt by new first-year birds, leading over time to gradual shifts in the local dialect. At the Edwin S. George Reserve, in Michigan, about 60% of Indigo Bunting songs were new in each year from 1979 to 1984. Most of these new songs failed to reappear in the following year, and fewer than 20% were copied by another male.

R. B. Payne estimated the half-life for the survival of songs to be 3·8 years. Survival of a song tradition over time is purely cultural. Because the birds do not learn their songs from relatives, and because females show no preference for males singing their father's song type, the persistence of a song type appears not to depend on any selective advantage that it may confer.

Cardinalids produce a wide variety of non-song vocalizations. According to Skutch, both sexes of the Black-faced Grosbeak give a series of 3-4, or sometimes 5-6, "chip" notes which may be a call or a song, but which differ from a more melodious song given only by the males. At least 13 calls are given by adult Lazuli Buntings, the most commonly used being essentially identical in structure to those of Indigo Buntings. Northern Cardinals have a vocabulary of at least 16 different calls. The calls of the last-named species, unlike its songs, appear not to be learnt; individuals raised in isolation have the same call repertoires as those of wild cardinals.

Non-song vocalizations of Northern Cardinals range from the faint sounds made by hatchlings to a variety of calls associated with specific aspects of adult behaviour. Loud "chip" calls may be used when a predator approaches the nest, or at similar moments of excitement. Female cardinals use the "chip" call primarily when approaching or leaving the nest. Males give a specific piping call during courtship feeding, and females make quiet "twitter" calls, possibly indicating submission, in the presence of their mate. Black-headed Grosbeaks have a "wheet!" call that they use only when taking flight. Young of this species, from a few days prior to fledging and for several weeks afterwards, give a "phee-oo" call when they sight a parent with food or hear it calling or singing. Dickcissels have a special call given on long flights, including migratory flights, that sounds like an electric buzzer.

The female Lazuli Bunting uses two calls to stimulate nestlings to beg. She gives a "took" call if the chicks do not beg when she arrives at the nest, and a "cheek" call if the "took" does not elicit the correct response. The begging calls of young Lazuli Buntings 2-5 days old are variable and non-specific, but by 7-8 days of age each chick within a nest has its own distinctive "signature" call. Male Indigo Buntings give a buzzy "tseep", audible for only a few metres, when approaching a female for copulation. This call, accompanied by song, is apparently necessary to arouse the female (see Breeding).

Although a number of cardinalids can make audiblenon-vocal sounds, including the whirring noise produced by the wings of a large flock of Dickcissels in flight, there seem to be no examples of specialized or deliberately made non-vocal sounds in the family.

Finally, turning to the two probable non-cardinalid generaParkerthraustes andSaltator (see Systematics), the Yellow-shouldered Grosbeak is vocally distinctive. Its thin, lisping song is quite unlike the songs of true cardinalids, but is reminiscent of that of some tanagers. The vocal behaviour of the saltators, however, appears never to have been studied in detail. Some saltators are notably fine singers, a characteristic that has made them, despite their sombre plumage, popular as cagebirds in South America (see Relationship with Man). The voices of the Black-headed, Lesser Antillean and Streaked Saltators, on the other hand, have been described as being harsh, and the call of the Black-headed has been said also to be "raucous". Songs of Greyish Saltator inhabiting dry and relatively open areas in northern South America are less tonal, and have wider bandwidths and higher frequencies, than those from more heavily vegetated habitats elsewhere within the species' range.Female song has been recorded for the Black-headed and Buff-throated Saltators.

Black-throated Saltators, which have a warbling song unlike that of other members of the genus, may sing togetheras a group during the early hours of the morning. Pairs of Greyish and Buff-throated Saltators sing antiphonal duets. A phrase of three or four syllables given by one individual is replied to with a similar phrase, and the process is repeated many times over. Skutch heard one female Buff-throated Saltator deliver both parts of this duet while approaching her nest. A longer and more powerful song is apparently given by the male alone, either from an elevated song perch or within cover. A rarely heard whisper song, given just before sunrise, is apparently addressed by the male to his mate.

Streaked Saltators generally sing from a lower perch thanthose preferred by the Buff-throated Saltator. Skutch described a flight display by the former, given at dusk, in which the bird rises in a circling flight to about 30 m into the air, singing loudly, before diving back into the shrubbery. Skutch never observed such a display performed by the three other saltators that he studied, namely the Buff-throated, Black-headed and Greyish Saltators. Black-throated Saltators sing in horizontal flight from one tree to another across the savanna, and Green-winged Saltators likewise sing in flight from tree to tree.

In the upper Térraba Valley, in Costa Rica, both Buff-throated and Streaked Saltators begin singing early in the year. They sing most persistently from the end of March until the beginning of the rainy season, in early June. Buff-throated Saltators also sing actively, mostly at dawn, during July and August, but they rarely do so from September onwards.

The songs of somesaltators vary geographically, a fact well known to Brazilian cagebird-fanciers.Skutch noted that Buff-throated Saltator song can vary considerably even between individuals living not many kilometres apart. Moreover, Black-throated Grosbeaks may not recognize the songs of conspecifics that sing in a different dialect.


Food and Feeding

Cardinalids are, by and large, opportunistic omnivores. Plant material, including fruits, flowers, buds, stems, seeds and occasionally nectar, normally represents a higher bulk in the diet than does animal matter. Black-backed Grosbeaks near Bogotá, in Colombia,have been recorded as foraging in the upper levels of arboloco trees (Polymnia pyramidalis), taking flowers and seeds and capturing insects from both flowers and the trunk. Northern Cardinals in Florida take nectar on occasion, and Yellow Grosbeaks in Mexico have been recorded as destroying flowers ofErythrina oliviae in order to rob them of nectar.

Some cardinalids are primarily arboreal foragers, while others, such as the Northern Cardinal and Painted Bunting, spend considerable time in feeding on the ground. Indigo Buntings, Dickcissels and other smaller species take seeds from grasses and herbaceous plants. A single species, the Indigo Bunting being a good example, may hunt for seeds and insects on the ground, explore fruiting shrubs for berries, and forage for insects high in the canopy.

In a recent worldwide survey of fruit-eating by birds, W. D. Kissling and his colleagues classified the Black-faced, Crimson-collared and Red-and-black Grosbeaks and the Orinocan Saltator as frugivores. Most other cardinalids are at least partially frugivorous; perhaps only the Dickcissel rarely, if ever, eats fruit, although there are no data on frugivory for the Vermilion Cardinal, nor for the Yellow-shouldered Grosbeak and some saltators.

In 1908, W. L. McAtee published a thorough survey of the stomach contents of species ofCardinalis andPheucticus native to the United States. For Northern Cardinals, he found wild grape (Vitis) to be the most important food item, the birds crushing the seeds as well as devouring the fruit. McAtee found that grape seeds constituted 17% of the Northern Cardinal's diet between November and April. Northern Cardinals in southern Ontario have been observed to crack and eat seeds of the riverbank grape (Vitis riparia).

Pyrrhuloxias eat the fruits of the christmasberry cholla (Opuntia lepticollis). Rose-breasted Grosbeaks in Pennsylvania are major consumers of elderberry (Sambucus) fruits, being second only to Scarlet Tanagers (Piranga olivacea) in this respect. Rose-breasted Grosbeaks are almost entirely frugivorous on migration, fruit amounting to 96% by volume of faecal samples taken at an autumn migration site on Block Island, in the US state of Rhode Island, although they take both fruits and insects on their wintering grounds in Costa Rica.

Yellow Grosbeaks observed in Jalisco,in western Mexico, commonly took fruits of the figFicus pertusa and of the hoop vine (Trichostigma octandrum), but visitedRecchia mexicana more rarely. Both Black-faced and Blue-black Grosbeaks at La Selva Field Station, in Costa Rica, have been recorded as taking fruits of Ficus pertusa. The Black-faced Grosbeak usually ate at least three fruits on each visit, with a mean 3·14 fruits over seven visits during two days, and it may be an important seed-disperser of this species. Fruit, however, appears insufficient to satisfy the protein requirements of these two species; T. C. Moermond and J. S. Denslow found that, unlike birds of purely frugivorous species, captive individuals of these grosbeaks, when fed on a fruit diet, lost weight if they were not given a protein supplement.

Fruit-eating cardinalids, like tanagers and emberizid finches, mash fruitsin the bill, separating larger seeds from the pulp and ejecting them before swallowing. Mashingbehaviour increases the processing time, but it allows fruit-eating birds to take larger fruits than those which they could comfortably swallow whole, and also avoids the inconvenience of having to carry large, indigestible seeds in the gut. Captive Black-faced Grosbeaks observed by D. J. Levey handled 3-5 small fruits at a time, crushing them for more than a minute and sucking the juices, before ejecting a dry mass of seeds from the bill.

Which seeds cardinalids include in their diet depends in part on their size and hardness. Even the heavy-billed Blue-black Grosbeak cannot handle the hard seeds ofHeliconia, but both it and the Black-faced Grosbeak can and do crack the smaller seeds ofPsychotria brachiata (Rubiaceae). The arillate seeds of a number of tropical plants attract cardinalids, as well as other bird species. Black-faced Grosbeaks, for example, eat the large red arils ofStemmadenia donnell-smithii (Apocynaceae). Skutch recorded this species as feeding on the arils ofAlchornea costaricensis (Euphorbiaceae), and Blue-black Grosbeaks as eating the oil-rich arillate seeds ofSiparuna nicaraguensis (Monimiaceae), plucking the seeds one by one from the pale red inner surface of the receptacle. Skutch also noted that wintering Rose-breasted Grosbeaks took the arils of five species of Costa Rican plant.

In Middle America,Blue-black Grosbeaks visit multi-crop fields known as "milpas" to feed on crops of maize (Zea mays), as well as visiting rice fields in wooded areas. Painted Buntings are primarily ground-foragers and seed-eaters. Analyses of the stomach contents of individuals from Florida and Texas showed that 70-73% of the diet consisted of seeds, including those of grasses and other plants. Lazuli Bunting diets are lower in seeds, which constituted only some 30% of plant matter in stomachs of this species in California.

Northern Cardinals are efficientin digesting seeds, although less so than some emberizid sparrows that specialize on seeds to a greater degree. According to a number of metabolic studies, cardinals are able to extract 75-84% or more of the utilizable energy from sunflower seeds (Helianthus), 85% from millet, 84% from sorghum (Sorghum), 79% from hemp (Cannabis) and 77% from ragweed (Ambrosia) seeds.

Cardinalids take varying amounts of animal prey, usually insects and other arthropods but, very occasionally, larger items. A Northern Cardinal, for example, has been seen to eat a mouse (Muridae). They may increase the proportion of animal prey in the diet during the nesting period. According to McAtee, animal matter normally amounts to some 30% of the diet of Northern Cardinals, but rises to as much as 78% in May, when the birds are nesting. Rose-breasted Grosbeaks also shift from a preponderance of plant material outside the breeding season to a diet of almost 75% animal matter while nesting. In north-eastern Venezuela, 83% of emetic stomach samples taken from Vermilion Cardinals contained the remains of invertebrates, while 48% contained fruits and 84% held seeds.

Blue Grosbeaks at five nests observed by F. J. Alsop in Tennessee fed their nestlings almost exclusively with large mantids(Mantidae) in excess of 75 mm in length, from which the adults had removed the head, wings and most or all of the legs. These insects accounted for more than 96% of items delivered to the chicks. Female grosbeaks, followed by their mates, plucked the mantids from weedy vegetation while in short or hovering flight, or while hopping on the ground.

Indigo andLazuli Buntings glean insects and spiders(Araneae) from the foliage of trees and shrubs, and Blue Buntings have been seen to peck at spider webs. In Hawaii, introduced Northern Cardinals hunt flying termites (Isoptera) in mid-air. Rose-breasted Grosbeaks are acrobatic feeders, capable of hanging upside-down from the ends of fine branches or pursuing flying insects on the wing. Individuals as young as 47 days have been seen to catch insects in the air. Females of this species hunting insects in New Hampshire foraged higher in trees than did the males, while males attacked prey on trunks and branches more often than did females, a pattern exhibited also, interestingly, by Scarlet Tanagers. Grosbeaks treat animal food much as they do fruits, mashing and tenderizing it in the bill before swallowing.

Yellow Grosbeaks are rare components of mixed foraging flocks in western Mexico. Farther south, on Cerro Pirre, in east Panama, flocks of 6-10 individuals of the localized subspeciessimulans of the Yellow-green Grosbeak regularly joined canopy flocks of New World warblers, tanagers, foliage-gleaners (Furnariidae) and woodcreepers (Dendrocolaptidae). They also foraged on their own, moving quickly through upper, outer branches, searching bark and foliage. Yellow-green Grosbeaks are common in mixed foraging flocks in Amazonian Brazil. Their congener the Black-faced Grosbeak associates in feeding assemblages that may consist of 30 or more individuals. Moermond has suggested that these large assemblages are the product of several "feeding units", presumably family groups, which join together (see Breeding). They have been recorded as searching for arthropods in curled-up dead leaves.

One of the few cardinalids with a considerably higher recorded proportion of animal material than plant matter in the overall diet, in McAtee's analysis 65% as against 34%, is theBlack-headed Grosbeak. This species and the Black-backed Oriole (Icterus abeillei) are the major avian predators of overwintering monarch butterflies (Danaus plexippus) in Mexico. Particularly on cooler days, when the insects are less active, the two bird species attack the butterflies in synchronous waves, acting as a single large mixed flock. L. S. Fink and L. P. Brower showed in experiments that the grosbeaks, unlike the orioles, appear immune to the emetic properties of the butterflies' cardenolide defences. Although the grosbeaks are eventually affected by the cardenolides, they are nonetheless able to consume the butterflies in large numbers, and may take more of the insects than the orioles do. The birds prefer male butterflies, which have lower cardenolide levels than do females, possibly distinguishing the sexes by taste during a preliminary "bite". Brower and W. Calvert have suggested that the pattern of synchronous attacks on the butterflies by grosbeaks and orioles may be maintained by cultural transmission from one generation to the next.

Dickcissels studied in Kansas, in the central United States, hunted through lower vegetation in search of insects or seeds, including grains of wheat. Throughout the breeding season they spent 17-21% of their time each day in foraging (see General Habits). Males devoted most of their time to feeding in forbs, where insects, particularly grasshoppers (Orthoptera), were more abundant than they were among grasses. R. L. Schartz and Zimmerman have estimated that Dickcissels need to eat 23 grasshoppers per hour to supply their energy requirements. There may be an increased peak of feeding activity in the evening if the temperature drops below 10-15·5°C, presumably to supply extra energy for temperature regulation during the night.

On the wintering grounds in Venezuela, Dickcissels formerly foraged on native grasslands, but they now gather in often enormous flocks that descend on local rice and sorghum fields, as well as feeding in fallow agricultural fields with abundant wild grasses such as Rottboellia cochinchinensis andOryza latifolia (see also Movements, and Status and Conservation). Although Dickcissels may prefer wild grains such asmillets (e.g.Echinochloa) over cultivated ones, rice and sorghum now make up some 80% of the birds' winter diet. Basili and Temple recorded that the birds fed in the morning and afternoon, but rested during the day. The longest feeding bouts involved feeding on immature green sorghum, one female eating 52 seeds in 241 seconds and the male 80 seeds in just under 290 seconds. Basili and Temple estimated that the 6,450,000 or so Dickcissels wintering in Venezuela eat 4,442,760 kg of rice seeds, 2,300,070 kg of sorghum seeds and 1,677,000 kg of wild grass seeds in the course of a single season, figures that nonetheless amount to no more than between 0·74% and 1·35% of the total crop.

Leaf-eating, folivory, is uncommon among small birds, presumably because of the lower ratio of nutritive value to bulk in a leaf diet. Indigo Buntings in Michigan eat leaves of aspen (Populus) and other trees on first arrival in spring, but not later in the season when other foods are more readily available. G. M. Sutton recorded Crimson-collared Grosbeaks in Mexico as "munching" leaves of shrubby nightshade (Solanum verbascifolium).

Folivory is much better documented for some saltators. Leaf-eating has been reported for Buff-throated and Black-headed Saltators in Costa Rica, Greyish Saltators in Costa Rica and Venezuela, Rufous-bellied Saltators in Bolivia, Masked Saltators in Colombia, Green-winged and Thick-billed Saltators in Brazil, and Orinocan Saltators in Venezuela. Moreover, E. S. Munson and W. D. Robinson have suggested that the Thick-billed Saltator, a primarily Brazilian species, may be an obligate folivore, the only one so far known among the Passeriformes. Leaves from ten species of woody plant, particularly the treePiptocarpha angustifolia (Asteraceae), accounted for 91% of the food items taken by saltators in the Marumbi State Park, in Paraná, during August, admittedly at the height of the dry season when fruits, which the birds perhaps eat at other times, are harder to come by. These saltators worked plucked leaves slowly back and forth through the bill, folding them lengthwise and pinching them until they formed small, mashed balls suitable for swallowing. The processing time was 5-26 seconds per leaf.

In thellanos of Venezuela, Greyish Saltators were found to feed almost exclusively on plant tissues, including a wide variety of fruits, leaves, flowers, flower buds, leaf buds, seedpod skins and tendrils from 49 plant species. Fruit and mature leaves made up, respectively, 36% and 28% of the diet. The birds defecated the intact seeds of fruits. During the dry season, when deciduous plants lost their leaves, the saltators ate a higher proportion of flowers and seedpod skins. When offered a choice, captives strongly preferred fruits over leaves or flowers, and wild leaves over cultivated leaves.

Leaves and flowers may provide birds with nutrients, such as calcium and phosphorus, that are scarce in fruits. Leaf-eating saltators, unlike larger obligate folivores such as the Hoatzin (Opisthocomus hoazin), appear not to retain food in the gut for long enough to ferment plant fibres. They appear to have adopted, instead, a "skimming" strategy, substituting rapid food intake and processing for more efficient digestion.Rodrí­guez-Ferraro and her colleagues found that captive Greyish Saltators retained leaves in the gut for an average of only 45 minutes, even less time than they did fruits, which were kept in the gut for a mean of 54 minutes. They found no evidence that the birds could break down cellulose fibres.Presumably, the birds satisfied their energy requirements, to the extent that they were able, from soluble materials in leaf cells. This may not be enough for an active life. Captives spent 50% of their time in resting, and, in a study by C. Bosque and his colleagues, both Greyish and Orinocan Saltators, which have a similar diet but take fewer leaves and a higher proportion of flowers, had lower basal metabolic rates than would be expected for birds of their size.

Saltators in general are mostly frugivorous, although they will also take nectar and may include arthropod prey in their diets. Golden-billed and Greyish Saltators in the Paraná Valley, in Argentina, consume a wide range of foods, plant seeds, particularly those ofSolanum amygdalifolium, predominating, but including also insects, primarily leafcutter ants of the genusAcromyrmex. The diets of the two species overlap broadly. Both appear to be opportunistic feeders, seasonal shifts in diet by the Golden-billed Saltator being due largely to the high availability ofSolanum seeds in the austral spring.

Streaked Saltators in Venezuela consume both fruits and seeds ofcolumnar cacti (Stenocereus). In Costa Rica, Buff-throated, Black-headed and Greyish Saltators feed primarily on fruits when not breeding, but switch almost exclusively to arthropods when feeding nestlings. Buff-throated and Greyish Saltators also eat the corollas of flowers, and Buff-throated and Streaked Saltators feed on the arillate seeds of a number of Costa Rican plants, includingProtium (Burseraceae).

Buff-throated Saltatorshave been seen to eat army ants (Ecitoninae), and ants have been recorded in their stomach contents. This was the third most commonly seen species attending army-ant swarms in Yacambú National Park, in Venezuela, in 2006.Rodrí­guez-Ferraro observedGreyish Saltators in Venezuela as they caught insects in July and September, when they were feeding chicks and moulting the body and wing feathers. Unusually, a mist-netted Black-headed Saltator in Costa Rica was found to have eaten part of a hummingbird (Trochilidae) adjacent to it in the net.

At least one member of the genus, theGreen-winged Saltator, is primarily an insect-eater. This species has been observed while feeding at a swarm of emerging winged termites, and, with Black-throated Grosbeaks, joining mixed-species foraging flocks inAtlantic rainforest.The Green-winged Saltator complements its diet with fruits, and has been seen to swallow fruits ofEugenia andSolanum granuloso-leprosum, and to gleanfruits ofPsychotria brasiliensis (Rubiaceae), crushing them with the beak turned upwards. Black-throated Saltators studied by C. Gomes de Almeida in Minas Gerais, in Brazil, foraged in small groups of, on average, three individuals, mostly on the ground or in low vegetation. Their diet consisted of 40% seeds, 36·9% flower buds and 13·8% fruit, as well as small numbers of invertebrates.

It has been suggested that the Rufous-bellied Saltator may be dependent on mistletoeberries of the genusTristerix, but this species has been recorded also as taking fruits of a number of other plants, includingBerberis andHeteromelas, as well as insects and seeds. An individual observed in Salta, in Argentina, fed mostly onBrachyotum berries, but also consumed young leaves of an introduced willow (Salix) and took young grass shoots from the ground.

The Masked Saltator may bea specialist on the cones ofPodocarpus oleifolius (see Status and Conservation). J. A. Tobias and R. S. R. Williams found this species to be a component of mixed flocks of tanagers and chlorophonias visitingPodocarpus stands in the Andes of southern Ecuador. The birds often lingered for long periods in the canopies of podocarps, feeding on the cones, and the observers never saw them take any other kind of food. L. M. Renjifo, who observed what may be a distinct population of the species in Colombia, did record it as taking other types of fruit, though in much smaller quantities. When eatingPodocarpus cones, which they do at all times of the year, the saltators concentrated on the seeds, discarding the fleshy red peduncle that attracted other birds.

The little-known Yellow-shouldered Grosbeak, the sole member of the genusParkerthraustes and, likeSaltator, probably not closely related to other members of the family (see Systematics), apparently forages as single individuals in the company of other canopy birds. It gleans items from foliage near the ends of branches.

So far as is known,"typical" cardinalids drink in the manner normal for most bird species, scooping up liquid in the bill and tilting the head upwards to swallow. Indigo Buntings drink only infrequently, but Varied Buntings do so occasionally, sometimes in groups, and Lazuli Buntings drink dew and rainwater from vegetation. Saltators are capable of suction-drinking, sucking liquid into the mouth, an ability otherwise unknown in the Cardinalidae, although Rose-breasted Grosbeaks have been seen to make throat movements while the bill is immersed in water. Suction-drinking is documented for many tanagers, among which it may have evolved as an adaptation for eating succulent fruits.



Little information is available on the mating system of Neotropical members of the family. Indeed, the breedingbehaviour of the Red-and-black Grosbeak, the sole member of the genus Periporphyrus, remains completely unknown. One Neotropical species, the Black-faced Grosbeak, may breed co-operatively, and the same presumably may apply to its congener, the Yellow-green Grosbeak, although the latter's breeding habits have not been studied. Skutch recorded that parent Black-faced Grosbeaks feeding the young were assisted by helpers, and Moermond observed groups of adults attending young in the nest at La Selva, in Costa Rica. Although the groups contained six to eight individuals, only four fed or defended the chicks. Moermond suggested that social organization in this species may consist of breeding a pair with one or two helpers, presumably offspring from earlier broods, accompanied by less closely related "hangers-on".

There is a single record of a Northern Cardinal nest shared by two females, one an albino, both apparently mated to a single male, and onerecord of a bigamous male in Ohio provisioning two independent nests and successfully fledging young from both. There are also at least two records, including a recent one from Oklahoma, of a single nest being occupied by two females, each mated to a different male. These appear to be incidents of nest-sharing, rather than co-operative breeding, although in the Oklahoma case one male fed all three nestlings.

All North American cardinalids with the exception of the Dickcissel are primarily monogamous, though with varying degrees of polygyny in some species. The Blue-black Grosbeak of Middle and South America, both sexes of which build the nest and feed the young, may likewise be monogamous. There is no evidence of any polygynousbehaviour, including extra-pair copulations, by Black-headed Grosbeaks and Varied Buntings. Only 3·7% of male Lazuli Buntings in western Montana had more than one mate at a time, and none changed its mate over the course of the season. Indigo Buntings have been characterized as polygynous, although most males have only one mate per season. At different sites in Michigan, 8·8-19·2% of breeding males had two, or even three or four mates, either simultaneously or in sequence. In Indiana, Indigo Buntings holding higher-quality territories, with high vegetation density and lush ground cover, were more likely to be polygynous. Polygynous males were just as successful at fledging young as were monogamous ones, suggesting that polygyny, if it can be achieved, confers an advantage. As males take only a limited role in caring for their young, females do just as well in terms of reproductive output when mated to a polygynous male as they do with a monogamous one.

Birds will occasionally mate outside their social pair, and females may settle on territory whether or not the male occupying it is mated. A female Indigo Bunting may make no effort to prevent her partner from mating with another, although the Payne husband-and-wife team recorded a case in which, after a new female Indigo Bunting mated with a returning male on his old territory, nested and laid eggs, the female of the previous year arrived and repeatedly chased the new one until she abandoned her nest. After her victory, the old female nested in her turn.

Dickcissels are polygynous, the male mating with up to six females, although males on poorer territories may go without a mate, and a variable proportion of males in a population will have only one partner. The number of females with which a male is likely to mate depends upon the quality of his territory, measured in terms of the number of possible nest-sites that it contains (see Habitat). E. J. Finck found that Dickcissels on prime territory in Kansas spent more time in territorial display than did those on less suitable areas, but the amount of time that they put in was not related to the number of females they managed to attract. Instead, the primary purpose of territorial display appears to be defence against other males. Competition for optimal territories appears to be intense. If a territory-holder in Finck's study area disappeared, it was replaced within a couple of days.

Pair-bonds of cardinalids may be short-term arrangements, or, as has been recorded for Northern Cardinals and Varied Buntings, they may be maintained for repeated nesting through the breeding season or even into succeeding years. Northern Cardinals usually join flocks in the winter months, but pair-members in Tennessee remain together and on territory throughout the year. Partners may leave wintering flocks together, but it is not known whether these are individuals that were mated in the previous season or whether they have paired up over the winter. Northern Cardinals have been observed to practise courtship feeding all year around, suggesting that, for this species, there may be some selective advantage in reinforcing the pair-bond outside the breeding season. It is possible that some Painted Buntings pair on their wintering grounds in Costa Rica, as partners have been seen together there. As males usually arrive on the breeding grounds in advance of females, it seems likelier that most pair formation takes place after arrival.

With Pyrrhuloxias the pair-bond tends to break down in late summer, but pair-members have been seen to be associating with each other in winter and it is possible that some bonds survive to the following season. Indigo Bunting mates separate after the completion of the breeding season, but roughly half of females whose males return in the next spring will reunite with their previous partner.The Payne couple has recorded Indigo Bunting partners that bred together for as long as five years, although this may be less a reflection of a long-term pair-bond than of the fact that each member of the pair returned to the same territory. The act of reuniting with an old mate did not improve or reduce a female's nesting success. Eleven percentof pairs broke up after the first breeding attempt, after which the female moved to a new territory and nested again with another male. She was more likely to do this if she had been successful in her first attempt, and she sometimes took her first brood with her to the new territory.

Dickcissels pair quickly once a female arrives on a male's territory. Similarly, female Black-headed Grosbeaks may select mates within hours of arriving on the breeding grounds, and it is possible that some pair before arrival, as may be true also for Rose-breasted Grosbeaks. Female Indigo Buntings, on the other hand, have been recorded as spending as long as 34 days on a male's territory before completing a nest, suggesting that pair formation can sometimes be protracted. M. Carey has suggested, however, that the birds may simply be postponing nesting because cowbirds are present in the area.

Pair-bonding itself is initiated with a series of courtship displays. Those of the Northern Cardinals include song flights in which the male, with the crest raised and the breast feathers fluffed, flies towards the female while singing, and alights near her. Aerial chases may follow. Males and females join in countersinging, and, in later stages, courtship feeding, during which the female assumes a submissive begging posture. Displays may show off carotenoid feather ornaments, including the breast feathers of males and the underwing-coverts of females, in the latter case an ornament that is normally not visible. These include lopsided displays in which the bird twists and angles its body to show off the feathers on its underside, including the underside of the tail, and often lifts the uppermost part of the wing away from the body.

North American cardinalids are particularly aggressive during the early part of the breeding season, when territories are being established. Males frequently chase each other in territorial disputes. Female Rose-breasted and Black-headed Grosbeaks are also aggressive towards one another during the breeding season. Male Lazuli Buntings are especially aggressive towards other males during pair-bonding, and also in the days preceding egg-laying, when females solicit copulation and their mates guard them. Chases can last for ten minutes as males pursue each other, flying through vegetation or spiralling into the air.

Male Dickcissels chase other males that fly across their territory, individuals taking up the pursuit as the flying intruder crosses one territory after another. Chases sometimes lead to physical combat as the birds grapple with each other, pecking at each other's eyes or pulling out feathers. Dickcissels encountering each other at territorial boundaries may drop to the ground and approach each other closely, flank to flank. Encounters end with a "winner" flying to a tall perch, while the "loser" squats in a submissive posture. Black-headed Grosbeaks and Painted Buntings have been recorded as fighting to the death with conspecifics. In contrast, there are no documented observations of Varied Buntings fighting one another, although this species will engage in mutual chases.

In addition to chases and fights, cardinalids express their aggression through an array of agonistic displays. These can include wing-flicking, tail-spreading and other actions. Pyrrhuloxias in territorial encounters raise the crest and fluff up the plumage. A similar display by Northern Cardinals is used for expressing fear or submission. Both Pyrrhuloxias and Northern Cardinals lower the crest before or during proper chases or attacks. Similarly, Rose-breasted Grosbeaks raise and ruffle the crown feathers, but are more likely to flee than to attack when in this posture. At higher intensity, when an attack is more likely, the feathers are sleeked.

During the breeding season, all North American cardinalids establish and defend territories that include both nesting and feeding areas. Whether this is invariably true in Middle and South America is not known, although Black-faced and, possibly, Yellow-green Grosbeaks are apparently not territorial. No cardinalid is known to be territorial outside the breeding season.

Males generally establish their territories early in the breeding season. Northern Cardinals may do so as early as January, and may defend territories for seven months. Male Pyrrhuloxias become aggressive as their winter flocks break up, while females remain sociable for longer. Lazuli Buntings establish their territories within a week of arrival on the breeding grounds, and all available territories are rapidly occupied during the first two weeks of the breeding season. A returning territory-holder will usually evict another male "squatting" on his territory, and young birds are unlikely to gain a territory until one of the older males fails to return. Early-arriving Painted Buntings settling on optimum territories are also likely to be evicted by former territory-holders once they arrive on the breeding grounds.

The measure of territory quality may vary from species to species. In the case of Dickcissels, quality is determined by vegetation profile (see Habitat), whereas for Painted Buntings on St Catherine's Island, in Georgia, the key factor may be the greater abundance of marsh grasshoppers in edge habitats, as opposed to interior-forest sites. In mountain valleys in central New Mexico, the key factor affecting territory quality for Black-headed Grosbeaks appears to be the numbers of Steller's Jays (Cyanocitta stelleri) and Western Scrub-jays (Aphelocoma californica), the chief predators of grosbeak nests. The highest-quality territories are on the valley floors, where jay density is lower. Yearling grosbeaks left to occupy territories on the valley slopes, which have a high density of nest predators, are unlikely to breed successfully. As they become older, Black-headed Grosbeaks shift from these lower-quality territories to higher-quality sites on the valley floor, and their nesting success improves accordingly (see Morphological Aspects).

Northern Cardinals, Black-headed Grosbeaks and Lazuli and Painted Buntings are known to return to roughly the same territories year after year. In southern Canada, some Northern Cardinals in London, in south Ontario, remained on the same territory for as long as five or six years. Almost all male Painted Buntings observed on St Catherine's Island returned to the same territory in the following year. In southern Michigan, 95% of older Indigo Buntings return to the same territory in successive years, but nearly half of those returning after their yearling season shift to a new territory. Yearling males, but not females, are more likely to disperse if they fail in their nesting attempts in their first season, but whether an individual remains on its old territory year after year or shifts to a new one seems to have little to do with its lifetime nesting success.

Female Dickcissels are not philopatric. On the other hand, of a population of males studied over a five-year period in Kansas, an average of 49% returned to the same territory in the following year. Dickcissels are polygynous, and males that had had fewer mates and fledged fewer young were more likely to switch territory than were more successful males.

Male cardinalids tend to direct their aggression towards other males of the species, although a male Varied Bunting may tolerate yearlings on its territory even if they are singing, attacking them only if they approach his mate. Similarly, female Northern Cardinals, Rose-breasted and Black-headed Grosbeaks and Lazuli Buntings are territorially aggressive towards other females. Compared with many passerines, both male and female cardinals are relatively high in testosterone levels throughout the year, and this may have some bearing on their territorial aggressiveness. Rose-breasted Grosbeak females are aggressive towards conspecifics of both sexes, although male-female encounters are more frequent than female-female ones, and agonisticbehaviour drops sharply after egg-laying begins. Female Pyrrhuloxias become defensive only if their eggs or nestlings are threatened.

Territory size varies a good deal, but for most species it tends to be, on average, roughly 1-1·5 ha. The size can differ considerably even within a species, depending upon such factors as territory quality and the density of other males. In general, territories tend to be smallest in the optimum areas, either because size is limited by competition from other males or because a smaller size can satisfy food and nesting requirements. Painted Bunting territories in Missouri had an average size of 1·44 ha if they were bordered by territories of other males, but isolated males held territories covering on average of 3·92 ha. In Texas, Pyrrhuloxia territories were smallest in areas where there were no Northern Cardinals, and were largest where Pyrrhuloxias were outnumbered by their congener.

Dickcissel territories in Kansas varied in size from week to week and over successive years. The average weekly territory size was 1·76 acres (0·7 ha) in 1966 and 1·25 acres (0·5 ha) in 1967. With this species, territory size depends on the density of males, which is, in turn, governed by the structure of the vegetation. Mated males have larger territories than do bachelors. In the central Platte River Valley, in adjacent Nebraska, Dickcissels begin nesting 2-3 weeks after Red-winged Blackbirds (Agelaius phoeniceus), and male blackbirds often chase Dickcissel males from their territories. The latter species' territorial defence diminishes as the season progresses. Males, particularly less successful ones, spend less time on their territories, an average of 7·5 weeks for males that subsequently switched territories, compared with 10·3 weeks for presumably successful males that returned to the same territory in the following season. Males may even abandon their territories before their females have completed their nesting duties.

Male Dickcissels periodically engage in flights away from their territories, particularly early in the breeding season. These distant flights may allow the birds to assess the numbers of males and the suitability of habitat in other areas (see Habitat). Unmated males will often abandon their territories and seek areas more likely to attract females; bachelor Dickcissels in Kansas usually remained on territory for less than two weeks

Courtship feeding is uncommonamong Rose-breasted Grosbeaks. Both this species and the Black-headed Grosbeak perform song flights that expose the black-and-white wing and tail feathers and the colourful wing-linings. H. R. Ivor documented a courtship dance of "indescribable beauty" by a captive Rose-breasted Grosbeak in which the male crouched on the floor of his cage, spread and quivered his wings, and sang with his head held tilted against the feathers of his back. The grosbeak followed this performance by lunging at the female and seizing her primaries, and copulation soon followed.

Dickcissels engage in sexual chasesthat end on the ground, where the male makes physical contact with the female and may pull her feathers. Male Painted Buntings, in addition to fluttering flight displays, display on the ground in a flattened posture, with wings and tail spread as they circle the female. Buntings performing this display reminded A. Sprunt of a "miniature turkey gobbler". Indigo Buntings sometimes perform a similar display. The male Indigo Buntings seldom sings during courtship, but he stays close to his prospective mate, chasing off other birds. At times he may force copulation on the female. A female Indigo Bunting will not solicit copulation unless she is in sight of a male, which must both sing and give the "tseep" call (see Voice). She will reply with a similar soft call, and the two will exchange further soft calls before mating.

The role that the bright male plumage of many cardinalids plays in female mate choice varies with species. For some, such as the Northern Cardinal, plumage has a direct role in mate selection. For others, including the Lazuli Bunting and the Dickcissel, quality of territory appears to be the deciding factor. E. Greene and his colleagues found that female Lazuli Buntings in Montana selected males on the basis of territory quality, rather than plumage brightness, preferring territories with high levels of shrub cover. Males holding these territories could be either very bright in plumage or very dull, and included yearling males able to settle in high-quality territories because they were tolerated by adults. Intermediate individuals had more difficulty in gaining good territories, and females were less likely to select them.

In terms of importance in mate attraction, the distinction between ornament and territory may not be clear-cut. Female Indigo Buntings appear to mate at random with regard to the genetic make-up of available males, and show no preference in respect of male song type (see Voice). Redder male Northern Cardinals may in themselves be more attractive to females, but they are also more likely to gain better territories as measured by vegetation density. Black-headed Grosbeak females apparently compete with other females for choice of territory, but active courtship by the male suggests that some direct choice of a mate may be involved.

The brilliant colour of male Northern Cardinals is unusual for a bird that is almost completely monogamous. Jawor and her colleagues have established, on the basis of intensity of red colour in a number of plumage ornaments and in the bill, that Northern Cardinals of both sexes mate assortatively, contradicting the results of an earlier study. Male cardinals, for example, prefer females with a brighter red bill and underwing-coverts. Each sex presumably prefers more brightly coloured mates, although it is possible that duller birds may be choosing similar mates by preference, rather than being left with less suitable partners after the brighter birds have paired off.

Different ornamentspossessed by cardinals may be correlated with, and thereby signal, different traits, and birds may be assessing several ornaments for various types of information before selecting a mate. Overall brightness, for example, correlates with body size, and bill colour with condition. Individuals with higher carotenoid loads have a better immune system-higher heterophil-to-lymphocyte ratios in males, and higher white blood-cell counts in females-possibly as a response to increased stress levels arising from aggressive encounters and increased parental care. Males with a brighter breast colour are better at provisioning young, as are females with brighter underwing-coverts. Even so, females pairing with brighter males feed the young less, presumably because the mate is doing a greater share of the work.

Cardinals continue to display their ornaments to each other even after pair formation. Because partners may divorce both within the season and during the winter, displays that reinforce existing pair-bonds may be important, especially as the birds nest repeatedly during the season. The female's underwing displays, which are performed away from the nest, may serve this purpose.

Females of blue-plumaged cardinalids may not rely directly on the structural blue and UV colours of males when selecting mates. B. Ballentine and Hill found that the intensity of male structural colours made little difference to female Blue Grosbeaks. Colour intensity may matter, however, in competition among males for better territories. Brighter male Blue Grosbeaks tend to be larger, and to have larger and higher-quality territories, but they are not necessarily better at feeding young-although they tend to feed first-brood nestlings at higher rates-suggesting that the qualities linked to colour intensity have more to do with competition among males than with parental ability. Bluer males, in addition to gaining better territories with greater food resources, are less likely to be cuckolded than are duller individuals. Although females with duller mates might be seeking extra-pair copulations with brighter birds, it is still not clear if structural colours are, in fact, true indicators of male quality.

Extra-pair copulations and the chief response to them, mate-guarding by the male, have been recordedfor the Northern Cardinal, the Blue Grosbeak, and the Indigo, Lazuli and Painted Buntings. Despite lengthy observation they have never been seen to be performed by Black-headed Grosbeaks, and Varied Bunting parters are so rarely out of each other's sight that such behaviour seems unlikely.

The frequency ofextra-pair copulations among Northern Cardinals is lower than that found for many other songbirds, producing 13·5% of nestlings in a Kentucky study by Ritchison. One reason for this may be vigilant mate-guarding by the male. Males maintained contact with their mates an average of 72·8% of the time during initial nesting attempts, and usually followed the female if she flew off. In one of the pairs that did raise an extra-pair nestling, the male spent only 19·4% of his time with the female. Both male and female Northern Cardinals make forays from the territory, presumably in search of extra-pair copulations, although it is possible that they may be seeking food or nesting material, and extra-pair young have been found in 14% of cardinal nests. Female cardinals provided with extra food actually made more trips off-territory than did unsupplemented females, supporting the hypothesis that extra-pair copulations, rather than foraging opportunities, may be the goal of such excursions.

In an Alabama study by L. K. Estep and her colleagues, 53% of Blue Grosbeak nestlings were sired by another male, and at least 70% of nests had at least one extra-pair chick. The intensity with which males guarded their mates depended on the brilliance, and presumably the attractiveness, of surrounding males, though it may be that straying females were attracted by the food supply on the other males' territories, rather than by their plumage. A similar situation prevails with Lazuli Buntings. The tolerating of dull males has its benefits for brighter birds; the females of the former are available for extra-pair copulations, and their presence in nearby territories serves as a "paternity buffer" against other bright males seeking extra-pair copulations themselves. In Greene's study of Lazuli Buntings, 49% of nests contained at least one chick not sired by the resident male.

Morphometric studies suggest that about 40% of the offspring of Indigo Buntings in Michigan were sired by a male other than the territorial resident, although only 2% of observed copulations were extra-pair. D. F. Westneat found through genetic examination that 14·4% of Michigan offspring had genotypes incompatible with one of their putative parents. In a later study in North Carolina, using DNA fingerprinting, he found that 35% of all nestlings came from extra-pair copulations, and that 48% of broods had at least one extra-pair offspring. In a few cases, the entire brood appeared to have been sired by another male. Westneat also found that males attempting extra-pair copulations were often unsuccessful, failing to make cloacal contact with the female, usually because the female resisted. Females were more likely to resist extra-pair mating attempts than to resist copulations with their mates. The male Indigo Bunting does not guard his mate so vigilantly as does the Northern Cardinal, but he may be more likely to copulate with her if an intruder is present.

Copulation by cardinalids appears to be typical of passerines in general. Reverse mounting, in which the female mounts the back of the male, has been recorded for a pair of Painted Buntings, the male in that case having been a yearling. Ivor's captive male Rose-breasted Grosbeak was also mounted by the female, but only after copulation. Although birds may copulate at various times throughout the breeding season, effective insemination by Northern Cardinals and Rose-breasted Grosbeaks may happen only during short periods prior to egg-laying. Indigo Buntings, with their more variable mating systems, appear to be capable of inseminating females for longer and more irregular periods.

Most north-temperate cardinalids begin nesting in early spring. In some areas Northern Cardinals may start as early as late February, while Rose-breasted Grosbeaks usually wait until mid-May. In the American Southwest, the laying of the first egg by Varied Buntings coincides with the first significant summer rainstorm. Blue Grosbeaks in the same region may delay the onset of breeding until the monsoonal summer rains, although farther south, in Costa Rica, females have been seen to gather nesting material in May, June and July.

In Costa Rica, the Blue-black Grosbeak reaches the peak of its breeding season in July and August, when the maize crops ripen.As with many other tropical birds, however, the nesting season may be protracted and can vary from place to place. In Colombia Blue-black Grosbeak nests have been found as early as February and as late as October, and in Brazil they have been discovered in March and April. The Glaucous-blue Grosbeak, in common with many south-temperate species, breeds between October and December in Uruguay. Black-backed Grosbeaks have been found nesting in November in Colombia, and Golden-bellied Grosbeaks in arid regions in western Ecuador nest during the rainy season, between February and May.

Some, but not all, members of the family are multiple-brooded. Northern Cardinals may begin a second brood as early as 13-14 days after the first brood fledges if no more than one young from the first attempt survives, but they will wait for an average of 19 days otherwise. Late clutches of this species can be very late indeed; in West Virginia, an active nest was found on 19th October. Pyrrhuloxias apparently do not raise a second brood, or they are rarely successful if they try, although a female has been seen while feeding fledged young as late as October. Vermilion Cardinals and Rose-breasted, Black-headed and Golden-bellied Grosbeaks are usually single-brooded, but in one Michigan case a pair of Rose-breasted Grosbeaks began a second clutch when the young of the first brood were only 2-6 days old.

Blue-black Grosbeaks may raise two or three broods in a season. Second broods are apparently common among Lazuli Buntings and Blue Grosbeaks, at least in the southern portions of their ranges. Varied Buntings, however, will normally renest only if their first breeding attempt fails before the eggs hatch. Dickcissels may renest if they lose their first brood, but usually they do not have enough time to rear a second before the autumn migration. There is a possibility that some very early nesters in the southern part of the Dickcissel's breeding range may move north and nest again later in the season.

Females of both the Indigo Bunting and the Lazuli Bunting normally take sole responsibility for nest-site selection and nest-building. Among other cardinalids, both sexes may help to select the nest-site or, as with Dickcissels, the female does so, her mate accompanying her. A female Northern Cardinal may visit a variety of locations,fluffing her feathers and turning around if she finds a possible site. The male follows her, and the two call to each other and, using the bill, manipulate nesting material. A female Indigo Bunting will often visit several sites before settling on one. She shows no apparent preference in subsequent nesting attempts for the plant species used for a previous successful brood. Instead, her selection appears to be based on such factors as the extent of canopy cover over the nest-site and the rigidity of the supporting stems.

Members of a pairof Rose-breasted Grosbeaks may turn around several times in an appropriate fork, as if testing its suitability as a nest-site. Pairs of Painted Buntings will investigate clumps of Spanish moss (Tillandsia usneoides) or other dense vegetation, searching through the foliage more rapidly than they do while foraging. When they locate a potentially suitable spot, either the female or, in 12 out of 32 observed cases, the male enters the vegetation and crouches. The female may arrange nearby foliage around herself.

Nest-sites are usually at low to medium heights among dense vegetation, usually under some sort of cover, although the only known nest of the Yellow-green Grosbeak was placed 7 m up, and in Canada the average height of 50 Rose-breasted Grosbeak nests near London, in Ontario, was 6 m. Cover may be essential for both concealment and thermoregulation. For Black-headed Grosbeaks, a primary concern in site selection may be that of keeping the nest sufficiently cool. Northern Cardinal nests can be as low as 0·25 m from the ground or as high as 12 m, the height above ground increasing over the course of the breeding season. In western Montana, 95% of Lazuli Bunting nests were within 1 m of the ground, with the remaining nests no more than 2 m up, although some nests of this species have been at 8 m. Indigo Bunting nests are frequently less than a metre from the ground; six nests followed in Michigan, for example, were at an average height of roughly 50 cm. Forty-five Painted Bunting nests found by D. F. Parmelee in Oklahoma were placed 0·3 m to 2·3 m above the ground, at an average height of almost 1 m. As with the Northern Cardinal, nests found late in the season were often higher up than those found earlier.

Dickcissel nests areoften placed near the ground, usually in dense vegetation in grasses, forbs or low woody plants with nearly complete overhead cover. Ninety-four nests found by T. G. Overmire in Oklahoma were at heights ranging from 8 cm to 4·25 m, measured from the ground to the top of the nest. The average height of 20 nests on the Arkansas Grand Prairie, where the birds optimally nested in briar patches bordering maturing oat fields, was approximately 1 m. A recent study by J. W. Walk and his colleagues of 505 Dickcissel nests in Illinois grasslands suggests that the birds will nest close to cropland, but may avoid nesting in areas of grassland close to a woodland edge.

Vermilion Cardinals may place their nests in a shrub or a prickly pear cactus (Opuntia). The Black-faced Grosbeak apparently conceals its nest among arboreal epiphytes. The Blue-black Grosbeak frequently builds its nest among low forest palms armed with a formidable array of spines. One such nest found by Skutch had been built on the remains of the nest of a Thrush-like Mourner (Schiffornis turdina), which had itself apparently been built over another grosbeak nest.

Subsequent broods are normally raised in a new nest, although a pair of Rose-breasted Grosbeaks used the same nest two years in succession, and there are, incidentally, two records of a Rose-breasted Grosbeak nest in Ontario being reused by a Wood ThrushHylocichla mustelina). Painted Buntings will sometimes reuse an old nest. Although Indigo Buntings rarely do so, one nest of this species was built by the constructing of additional layers over an older one. Lazuli Buntings have not been recorded as reusing a nest, but females may return to the same bush, and occasionally the same location within it, as that selected in the previous year. Sutton described an active Dickcissel nest that had been built on top of a failed nest used earlier in the season.

Both sexes ofthe Blue-black Grosbeak take part in the building process, bringing nesting material and sitting in the nest to give it shape. Both sing as they build.Both themale and the femaleof the Rose-breasted Grosbeak build, whereas the female of the Black-headed Grosbeak does most of the work of nest construction. With most members of the family, the female builds a nest either alone or, as in the case of the Northern Cardinal and perhaps the Blue Grosbeak, with only limited contributions from the male. The male may accompany his mate during the process, but he is more likely to stay nearby and sing as she proceeds.The nest of the Black-faced Grosbeak is built by both partners, although the female appears to take the major role. J. K. Strecker reported in 1893 that both sexes of the Painted Bunting take part in nest-building, but more recent observers state that the female builds alone.

Northern Cardinals and Black-headed Grosbeaks carry out most of the nest-building work in the early hours of the morning, but female Lazuli Buntings build at any time of day. Female Painted Buntings in Oklahoma built after sunrise and sunset, but those attending an earlier brood while at the same time constructing a second nest built at sporadic intervals throughout the day.Dickcissel males may interrupt their nest-building mates for bouts of courtship and copulation.

The building of thefirst nest of the season can take much longer than the time required for subsequent ones. Northern Cardinals may spend 2-3 weeks on their first nest, while Black-headed Grosbeaks take 3-4 days. Lazuli Buntings may take 5-7 days to build a nest early in the season, but they can complete a later nest in as few as two. A female Dickcissel will begin the construction of a number of tentative nests after she arrives on the breeding grounds, but she soon focuses on a single site and completes the building work in about 2-4 days.

So far as is known, all members of the family build open cup-nests. There is, however, a single record of a Blue Grosbeak pair in Arkansas that nested, apparently successfully, in an enclosed nestbox designed for Eastern Bluebirds (Sialia sialis). Nests are constructed from twigs, grasses or plant fibres, and lined with finer material. A Black-headed Grosbeak has been observed to steal material from a recently deserted nest of a Brewer's Blackbird (Euphagus cyanocephalus). Indigo Buntings bind the nesting materials in place by weaving and wrapping them with spider web, especially around the rim, and incorporate dead leaves into the nest structure. Northern Cardinals add bits of paper such as chewing-gum wrappers, and Indigo Buntings and Blue Grosbeaks tend to include pieces of paper, shed snakeskin and similar items.Blue Grosbeak nests frequently contain pieces of cellophane, perhaps incorporated as a substitute for snakeskin.

Northern Cardinals and Pyrrhuloxiasdo not attach the nest to the surrounding branches, but simply wedge it into place.Passerina buntings anchor their nests in place by weaving strips of grass or bark around supporting twigs. Dickcissels either do not bind their nests to surrounding vegetation or make only a flimsy linkage, relying instead on a number of supporting plants to hold the nest.

The nests of Rose-breasted Grosbeaks can be so sparsely built that the eggs can be seen through the base from below, but they may contain slender forked stems of plants such asGeum canadense to provide structural strength without adding bulk.The Black-faced Grosbeak's nest is a flimsy structure, with an unusual lining of green bromeliad leaves. Blue-black Grosbeaks may line the nest with black fungal filaments, while Varied Buntings in Arizona line the nest with reddish-brown to yellow stems and the inflorescences of littleseed muhly (Muhlenbergia microsperma). They may delay nesting until there has been enough rainfall to allow suitable plants to grow.

Nests can become infested with mites (Acarina) and other invertebrates. Northern Cardinal nests in Ontario each yielded hundreds or even thousands of mites, mostlyOrnithonyssus sylviarum, as well as spiders, bark lice (Psocoptera) and insects of various sorts. A springtail,Entomobrya assuta, may have a particular affinity for cardinal nests.

Egg-laying begins only a few days after the nesthas been completed, fewer than six days later in the case of Pyrrhuloxias but after only a day or two in Dickcissels. Black-headed Grosbeaks typically begin to lay eggs 2-3 days after nest-building ends. So far as is known, members of the family lay eggs sequentially, one per day, until the clutch is complete. Laying normally takes place early in the morning, within an hour after sunrise in the case of Black-headed Grosbeaks and Indigo Buntings, but as late as three hours after sunrise in the Rose-breasted Grosbeak.

In general, eggs of cardinalids range in colour from whitish to blue or greenish-blue, usually with some degree of brown, black or purplish mottling, spotting or streaking, particularly around the blunt end. Blue-black Grosbeak eggs, for example, are either white or tinged with blue, blotched and marked with chestnut, reddish-brown or pale lilac. Indigo Buntings, Dickcissels and some other species normally lay unmarked eggs. Some Indigo Bunting clutches contain both unmarked and spotted eggs, but there is no evidence to suggest that these are laid by different females.

Members of the family breeding in temperate climates normally lay three or four eggs in a clutch. Dickcissels may lay as many as six, although a proportion of the eggs may be sterile, possibly as an adaptive response to the ability of the female to provision a large brood. Dickcissel second-brood clutches on the Kentucky-Tennessee border contained, on average, about one egg fewer than did first broods. Clutch size of tropical species is generally lower, between two and three eggs. One Black-faced Grosbeak clutch contained three dull white eggs, mottled and spotted with brown. The Blue-black Grosbeak lays two eggs, a typical clutch for a tropical passerine.

The clutch size of Northern Cardinalsin Bermuda is significantly lower, by 18·2%, than that of the species in coastal North American habitats at similar latitudes. As cardinals have been in Bermuda for only a few centuries, Crowell and S. I. Rothstein believe that this probably represents not an evolutionary change, but a fine-tuned response to proximate factors in the environment, including a more stable climate.

Among most cardinalids for which the relevant information is available, incubation is carried out by the female alone, as typified by, for example, the Black-faced and Blue-black Grosbeaks, although the male usually stays in the vicinity.Incubation normally does not begin until the last egg or the penultimate one is laid, although Rose-breasted Grosbeaks may start to sit as early as the second day of laying, and female Northern Cardinals sometimes sleep on the nest after depositing the second of a three-egg clutch.

Males of theIndigo Bunting and the Dickcissel ignore the female during incubation. On the other hand, male Northern Cardinals, Pyrrhuloxias and Blue Grosbeaks may bring food to the female as she sits on the nest. When the female Black-faced Grosbeak returns to the nest, the male accompanies her in a looping flight as both birds sing. Skutch once recorded a female accompanied in this fashion by two other individuals, one presumably a helper (see above).

Males of the Rose-breasted, Black-headed and Golden-bellied Grosbeaks undertake some of the task of incubating the eggs, but the role of the male of otherPheucticus species is apparently not known. In the case of the Rose-breasted Grosbeak, the male may be responsible for one-third of the incubation work during the day, and male Black-headed Grosbeaks have been recorded as contributing up to 50% of the daytime incubation. Male grosbeaks may develop bare areas of skin similar to the brood patch of the female, but less heavily vascularized. Females apparently do all of the incubation at night.

Female Indigo Buntings spend up to 79% of the daytime hours in incubating. The female Blue-black Grosbeak stays on the nest for approximately 70-75% of the time, but may incubate for as much as 95% of the day in rainy weather. One Blue-black Grosbeak observed by Skutch on a cool rainy day spent six hours in continuous incubation, the longest period that he recorded for any passerine bird. At two of three nests that Skutch observed, the male occasionally brought food to the female while she sat on the eggs.

The eggs generally hatch after 11-14 days, 12 days being the average for most cardinalid species. Young Northern Cardinals may take from 12 to 24 hours to break out of the egg, whereas Dickcissels can free themselves from their shell within a few minutes. The entire brood usually hatches over the course of one or two days. The young usually emerge unaided, but occasionally a female Northern Cardinal will lend assistance. Hatchling cardinalids are sparsely covered with down, generally either grey or white in colour. The colour of the mouth-lining ranges from orange to deep red, with yellow flanges or edging at the gape. Lazuli Bunting hatchlings retain the egg tooth for up to three days. The eyes of hatchlings are initially closed; they may open by three days in Painted Buntings, by five or six days in Varied Buntings, and by as late as eight days in Black-headed Grosbeaks. Black-headed Grosbeaks grow in weight from 3 g to 30 g during their time in the nest.

With many species, including the Black-faced Grosbeak, only the female broods the nestlings, but males of the Rose-breasted and Black-headed Grosbeaks brood occasionally, the Rose-breasted Grosbeak doing up to 25% of the daytime brooding. Brooding sessions are often short; in 1915, I. N. Gabrielson noted that brooding stints of a nesting Rose-breasted Grosbeak lasted for an average of seven minutes each, stretching to 17 minutes in the last two days, when the female brooded not by sitting, but by standing on the nest with wings slightly spread. Brooding of chicks is most continuous in the first few days after hatching, before the young are able to control their own body temperatures. Northern Cardinals are able to thermoregulate by five days of age, and Dickcissels by six days. A female Dickcissel watched by A. O. Gross brooded her young closely for the first three days, when ambient temperatures rose to above 32°C, but she was less attentive on days three and four and did not brood after day six. A six-day-old nestling left alone in the nest for one hour maintained an internal temperature of 41°C, even though the ambient temperature was only 26·6°C.

Nestlings are usually fed mostlywith animal matter, including grasshoppers and caterpillars. Pyrrhuloxias may feed their young with cactus fruit, as well as insects. Rose-breasted Grosbeaks include a small percentage of seeds, which they may shell at the nest before passing the kernels to their young. Female Dickcissels, which tend their young alone, may start feeding as little as 30 minutes after a chick hatches. One was seen to bring grasshoppers to a five-day-old brood of four chicks at the rate of 20 per hour.

Both sexesof Cardinalis and of North American species ofPheucticus feed young in the nest. Male Northern Cardinals studied in Ohio were, unusually for such a highly dichromatic species, assiduous in tending their young, adjusting their efforts to the size of the brood to an even greater degree than did females, and increasing their feeding rates as nestlings aged. T. S. Filliater and Breitwisch explored a number of potential explanations for this, including the possibility that females may threaten their mates with divorce if they do not "measure up". A later study of the same population found that, on the basis of reactions to models of predators, males contributed equally with females to nest defence during the middle of the nesting period, although defence seemed to be only a minor factor in nesting success. Males of the Vermilion Cardinal, another highly dimorphic species, are similarly attentive to their young.

Among Passerina species, the female normally takes responsibility for most or all of the care of the nestlings, including feeding. Male Lazuli Buntings are highly variable in their contribution to nestling welfare, but in one Montana study males were responsible for 24·6% of feeding visits to the nest. Male Painted Buntings apparently do not feed their nestlings, and Varied Bunting males do so only after the fourth or fifth day. Indigo Bunting males usually do not feed the chicks directly, but may do so to a limited extent nearer the time when the young leave the nest, and may pass food to the female, she then feeding it to the chicks. Monogamous male Indigo Buntings are more likely to feed their young than are polygynous ones.

Although both sexes of the Blue Grosbeaks observed by Alsop in Tennessee fed their young, most feeding and all of the prey capture, involving mostly large mantids (see Food and Feeding), was done by the female. Keyser and Hill found that female Blue Grosbeaks fed their nestlings more than six times more frequently than did males. Male Blue Grosbeaks may feed young from a first brood while the female is building a nest for the second, as male Rose-breasted Grosbeaks will do on the rare occasions when a female renests. Similarly, male Painted Buntings will take over the care of an earlier brood while the female tends to a later one. Blue-black Grosbeak nestlings are fed by both parents, and Black-faced Grosbeak nestlings are provisioned both by parents and by helpers (see above). H. Brackbill recorded a juvenile Northern Cardinal that fed nestlings raised by its presumed parents; the juvenile also begged from the adults, and the male did feed it, though the female attempted to drive it away.

Dickcissel malesnormally take no part in the feeding of the young. Gross recorded a male that failed to feed his young even after the death of the female parent; the brood starved to death. The male's contribution, however, is apparently variable. In a population of Dickcissels studied in east-central Illinois, J. D. Maddox and E. K. Bollinger recorded males as feeding young at six out of eight nests during 1997, males accounting for 37% of total visits to all nests. This unusual behaviour may have been prompted by a scarcity of food, although this was not assessed. In the following year, when food was presumably more abundant, no males were recorded as visiting nests.

Except for species such as the Dickcissel, both sexes generally carry away the nestlings' faecal sacs, although in the first days after hatching the adults swallow the droppings, rather than remove them. Lazuli Bunting females have been seen to insert the bill deep into the bottom of the nest and to shake the nest vigorously for 10-15 seconds. The function of this behaviour is unknown, but it may have the effect of shaking ectoparasites lower into the nest.

Although male Indigo Buntings often leave most of the nesting duties to the female, they may help to defend the nest. Black-headed Grosbeaks react to potential threats by flying from branch to branch, uttering loud distress calls. As the young in the nest grow older, and thereby represent a greater cumulative parental investment, the calls begin sooner, and increase in number, at the first approach of a threat. Similarly, Westneat found that Indigo Buntings give many more alarm calls per minute when disturbed during the nestling phase than when a potential threat appears during incubation.

Young cardinalids generally leave the nest at between 9 and 12 days of age. Some exceptions to this are presented by Northern Cardinals, which sometimes depart at only 7 days, especially if disturbed, and Black-faced Grosbeaks, which have been recorded as remaining in the nest for 15 days. Blue-black Grosbeak nestlings are fully feathered by 10 days, and leave the nest after 11 or 12 days, as do most Black-faced Grosbeaks.

After having left the nest, the young may move very little for the first few days, up to eleven days in the case of the Northern Cardinal, but they gradually acquire more mobility and independence. Juvenile Pyrrhuloxias start to form flocks as early as July, and young Blue Grosbeaks form flocks that are later joined by parent birds and the offspring from later broods. Male Black-headed Grosbeaks sometimes leave the breeding area before either females or young, and the last young of the year often do not depart until several weeks after the last adults.

Parents continue to feed their young for several weeks after fledging, Lazuli Buntings and some Dickcissels doing so for two weeks and some other species for longer, and the young may continue to associate with their parents for some time after they are able to feed themselves. Rose-breasted Grosbeaks in southern Ontario are dependent on their parents for three weeks after leaving the nest, before they disperse from the natal area. One juvenile Northern Cardinal in Maryland, in the east United States, was independent by 50 days, but it associated with its parents for a further 20 days. Dickcissel fledglings studied by K. M. S. Wells and her colleagues in south-western Missouri were independent after two weeks, but they moved little until 28 days after fledging. Some Dickcissel young may be dependent on their parents for four weeks or more after leaving the nest. Indeed, one parent observed by Walk and his colleagues in south-eastern Illinois continued post-fledging care for 43 days. This long period of post-fledging care may be a factor explaining why most Dickcissels are single-brooded. Male Painted Buntings observed by Parmelee attended their young fledglings, but normally did not feed them unless the female was engaged in constructing another nest. One male took over care of his brood six days after they had left the nest, while the female was renesting. Parmelee has suggested that the ability of polygynous male Painted Buntings to care for fledged broods may be the critical factor determining whether a female renests.

Some odd associations between parent Northern Cardinals and other species have been recorded. A pair of cardinals in Maryland shared a nest with a pair of Song Sparrows (Melospiza melodia); both species fed the young and cleaned the nest. A Canyon Towhee (Pyrgisoma fusca) has been recorded as actively feeding fledged cardinals for a period of three weeks; the true parents were in attendance and accepted the assistance. In contrast, a male Northern Cardinal was observed when feeding a brood of fledgling American Robins (Turdus migratorius) over a period of several days while provisioning his own young, and another male was recorded as tending a nest containing three young Yellow-breasted Chats (Icteria virens). In another interesting case, a female cardinal in Iowa shared a nest with an American Robin, the nest containing three robin eggs and two cardinal eggs. The two females, attended by their respective mates, incubated the eggs separately or together, the two sitting side by side on the nest, although only the robin chicks fledged successfully. As the female cardinal became increasingly aggressive to the female robin over time, it may be that thisbehaviour was the outcome of competition for a suitable nest-site. Finally, there are recent records of fledgling Carolina Wrens (Thryothorus ludovicianus) taking refuge in a newly lined cardinal nest during inclement weather, and of a failed cardinal nest being used by a Grey Catbird in the same season, but there was apparently no direct interaction between the two species in either case.

Conspecific brood parasitism issuspected in less than 1% of Indigo Bunting nests in Michigan, but is otherwise rare or absent in this family. There is one recent record of the appearance of three eggs in a Dickcissel nest, already containing four, during a two-day period. As Dickcissels lay only one egg per day, one of the eggs must have been laid by another female.

Every cardinalid species breeding north of Mexico is aknown host of the Brown-headed Cowbird. In addition, Bronzed Cowbirds (Molothrus aeneus) have been recorded as parasitizing Northern Cardinals, Pyrrhuloxias and Painted Buntings. Shiny Cowbirds (Molothrus bonariensis) are known to parasitize Golden-bellied Grosbeaks and some saltators (see below). Cowbirds may be particularly attracted to Northern Cardinals because the latter's bright colours and conspicuous behaviour make their nests easier to find. A high density of cardinals in the area was the most important factor influencing the rate of cowbird parasitism of the Black-capped Vireos (Vireo atricapilla) in Texas. Despite this, young cowbirds often do not survive well in cardinal nests. Cardinals studied by D. M. Scott and Lemon at London, in Ontario, produced only about 25% of the number of young cowbirds successfully raised by other host species, particularly in mixed broods, in which the young cowbirds were often underweight and grew poorly. Unlike some other hosts, Northern Cardinals have an incubation period only slightly longer than that of cowbirds, so that parasite eggs laid after incubation has begun are less likely to hatch.

The frequency with whichcowbirds parasitize members of this family varies among both species and populations. The Indigo Bunting is one of the 17 most frequent cowbird hosts throughout its range, and one of the seven most heavily parasitized species in the Atlantic coastal plain and Piedmont regions of South Carolina and Georgia. Similarly, as many as 71-100% of Lazuli Bunting nests were parasitized by cowbirds in the Sacramento Valley, in California, as were at least 50% of Blue Grosbeak nests in Grand Canyon National Park, in Arizona. Rose-breasted and Black-headed Grosbeaks are rarely parasitized, usually fewer than 10% of nests being affected. Aggressivebehaviour and nest-guarding by these host species may keep parasitism rates low. Nevertheless, a rate of 23% of nests has been recorded as being parasitized in the zone of range overlap between the two grosbeak species, and one Black-headed Grosbeak population in British Columbia, in south-west Canada, had a parasitism rate of 35%.

Dickcissels and Bell's Vireos (Vireo bellii) were the two species most heavily parasitized by Brown-headed Cowbirds at the Konza Prairie Biological Station, in north-east Kansas, between 2002 and 2007. Their rates of parasitism, 69·6% for the Dickcissel and 70·5% for the vireo, were considerably higher than those for the next most heavily parasitized species, the Red-winged Blackbird, for which the rate was 21·9%. Some other species on the station, including the Blue Grosbeak, had 100% of their nests parasitized, but the number of nests involved, only four in the case of the grosbeak, was much smaller. On the Konza Prairie the Dickcissel appeared to be the cowbird's preferred host. A Dickcissel can fledge up to five cowbirds from a single nest, and, unlike the vireo, it rarely abandons a nest with cowbird eggs in it. In terms of the number of young cowbirds that these nests produced, Dickcissels led the way, with 1295 fledged offspring from 619 parasitized nests over the course of the study. These parasitism rates were in fact lower than the figure of 84·7% calculated by Zimmerman for the same site from data collected between 1965 and 1979.

Cowbird parasitism rates can range from zero to 100% for one and the same species. In the case of Lazuli Buntings, nests situated close to cowbird roosts may suffer very high parasitism, while more distant ones experience none. Yearling birds may experience higher rates of parasitism than older ones, perhaps because older birds are more likely to occupy territories with greater opportunities for nest concealment. Among Dickcissels, parasitism rates range from non-existent, as in prairie barrens on the Kentucky-Tennessee border, to 100%, with the heaviest rates in the Great Plains. The reasons for this variation are not altogether clear. Rates have been reported to be highest in areas where nest density is low, and vice versa, but this was not the case in the Flint Hills of Kansas and Oklahoma, where cowbird parasitism, in a study by W. E. Jensen and J. F. Cully, varied from zero to 92% along a latitudinal gradient from south to north on ungrazed prairie, with no apparent correlation between parasitism and habitat features. In an earlier study in the same region, however, the Dickcissels experienced their highest levels of brood parasitism in unburnt hay fields, where nest density was about half that on restored prairie grasslands (see Habitat).

Habitat may not always be a factor in determining the risk of cowbird parasitism (but see Habitat). For example, Indigo Buntings in West Virginia were more likely to be parasitized by cowbirds if the vicinity of the nest had lower cover and a greater number of tall snags, but for those in Michigan the habitat, including the choice of nest-site, seemed to make no difference.

Parasitism rates canvary through the season, making estimates based on short-term studies unreliable as general guides. Parasitism of Indigo Bunting nests decreases through the late spring and summer, because cowbirds stop laying before the buntings embark on late-season clutches. At the Edwin S. George Reserve, in Michigan, the Payne couple found that, while parasitism rates were 23% in May and 26% in June, they declined to 9% in July and reached zero in August. At Niles, also in Michigan, more than half of Indigo Bunting clutches were begun after the peak of the cowbird season had passed.

Lazuli and Indigo Buntings will desert a nest if a cowbird lays in it first, whereas a Northern Cardinal in the same situation may not desert. A Blue Grosbeak will reportedly build a new nest on top of its old one if a cowbird has laid eggs in it. In many cases, however, birds will simply accept cowbird eggs, and the reproductive cost of brood parasitism may be slight.

The presence of cowbird chicks seems to have no effect on the ability of Northern Cardinals and Rose-breasted Grosbeaks to raise their own young. K. P. Eckerle and Breitwisch, who studied the effects of parasitism on Northern Cardinals in Ohio, observed adults taking action against cowbirds only at three parasitized nests, 5% of the total. Two abandoned the nest without laying, and a third buried the cowbird egg under new nesting material.

Other species do not fare so well if cowbirds lay in their nests. Cowbirds may remove eggs of the hosts before laying in the nests of Dickcissels, Indigo Buntings and other species, although eggs removed from Northern Cardinal nests may have been taken out by other cowbirds. Cowbirds may also break eggs by pecking at them. It has been suggested that an unusual statistical difference in eggshell mass between western and eastern populations of Painted Buntings, western females laying eggs with heavier and presumably less fragile shells, may be a response to the historical presence of egg-pecking cowbirds in the range of the western birds, and their absence, until comparatively recently, among the eastern population (see Status and Conservation). Pyrrhuloxias in Texas fledge very few young if they have been parasitized by cowbirds; in one parasitized brood, two Pyrrhuloxia chicks were found trampled to death in the bottom of the nest. Parasitized Dickcissel nests also reportedly fledge fewer young than do unparasitized ones. Cowbird-removal experiments in the Flint Hills nonetheless made no difference to nest productivity, despite a parasitism rate of more than 80%, with multiple cowbird eggs per nest. This may have been because the cost of parasitism was swamped by the much greater effects of nest predation, which was responsible for the overwhelming majority of nest failures.

M. A. Whitehead, S. H. Schweitzer and W. Post monitored nests of Northern Cardinals, Blue Grosbeaks and Indigo and Painted Buntings on James Island, in South Carolina, in order to determine the effect of cowbird parasitism on nesting success. Parasitism reduced the mean clutch size, but not the hatching rate, for all four species, and reduced the fledging rate for Blue Grosbeaks. It had little effect overall on the seasonal fecundity of the hosts, which were still able to raise young successfully. Indigo Bunting nests parasitized by cowbirds may be less likely to fledge young: in a survey of undisturbed nests in which at least one egg was laid, 56·4% of 1310 unparasitized nests were successful in fledging at least some nestlings, but only 19·5% of 411 parasitized nests did so.

Egg removal and loss probably represent the most serious cost of parasitism to Indigo Buntings, but the presence of cowbird young imposes costs of its own. Indigo Bunting nests with only one cowbird egg were more likely to produce young than were those with two or more, 22·1% of 339 nests with one egg being successful, compared with only 6·9% of 72 nests containing two or more cowbird eggs. Young cowbirds hatch a day or two earlier and grow more quickly than young buntings. They outcompete and overcrowd their smaller nestmates, and may even push them out of the nest. In captive mixed broods studied by R. B. Payne, the nestling cowbirds trampled the nestling buntings, climbed on them and pecked them about the head. A nestling cowbird has been videotaped as it ejected an Indigo Bunting chick from the nest, but it is not known how common such behaviour is. Young cowbirds even affect the survival of Indigo Bunting chicks after they have left the nest, presumably by competing for the attention of their foster parents.

The presence of cowbirdscan have an adverse effect on general health and fitness of the host's young. D. C. Dearborn and his colleagues found that Indigo Bunting chicks in parasitized nests received less food, begged more, gained mass at a slower rate, and were generally in poorer condition than nestlings from cowbird-free broods. Parasitized nests are more likely to succumb to predators, possibly because adults are forced to make more provisioning visits to the nest and can therefore spend less time in guarding their brood. Only 17% of young Indigo Buntings fledged from parasitized nests in Michigan returned in the following year, compared with 9·3% of young raised without cowbirds.

Predation is the single most important reason for nest failure, as it is for most passerine birds. While parasitism on James Island was directly responsible for only 2·9% of nest failures for four cardinalids and two other passerines, predation by other birds and reptiles caused 84·1% of nest losses, with additional failures due to predation by mammals and fire ants (Solenopsis). Nest predation by birds, snakes or mammals, including squirrels (Sciuridae) and northern raccoons (Procyon lotor), is the primary cause of nest failure among Rose-breasted and Black-headed Grosbeaks. Rose-breasted Grosbeaks reacted to a nest-raiding American red squirrel (Tamiasciurus hudsonicus) with a series of screams and threat displays that involved either tilting the head upwards and ruffling the wings or spreading and waving the wings, presumably to expose the colourful underwing-coverts. C. P. Ortega and J. C. Ortega identified the main nest predators in Black-headed Grosbeak habitat in riparian and oak pastures in Colorado as Black-billed Magpies (Pica hudsonia), Western Scrub-jays, Steller's Jay, rock squirrels (Spermophilus variegatus), chipmunks (Tamias), deer mice (Peromyscus maniculatus), long-tailed weasels (Mustela frenata), striped skunks (Mephitis mephitis), northern raccoons, and garter snakes (Thamnophis).

Rose-breasted Grosbeaks in southern Ontario woodlots suffered a 63% predation rate, 95 of 151 nests being lost. Black-headed Grosbeak nests studied by the Ortega pair in Colorado had a much better success rate, only 37% of 41 nests entirely failing owing to predation. Indeed, the grosbeaks had the highest success rate of passerines in the area, even though their nests, which were sometimes decorated with the purple flowers of lupines (Lupinus), appeared conspicuous to human observers. Their high success rate may have been due to aggressiveness in nest defence by both parents.

Non-venomous snakes known as racers (Coluber constrictor) have been seen to swallow a clutch of Dickcissel eggs in the Flint Hills, in Kansas, and to attempt to seize a nestling Indigo Bunting in Missouri. Blue Jays (Cyanocitta cristata) and Common Grackles (Quiscalus quiscula) kill young Rose-breasted Grosbeaks. Steller's Jays and Western Scrub-jays are the chief predators of Black-headed Grosbeak nests, but gopher snakes (Pituophis catenifer) also take this species' eggs and squirrels take its young. Plush-crested Jays(Cyanocorax chrysops) frequently raid the nests of Black-backed Grosbeaks. A Barred Forest-falcon (Micrastur ruficollis) has been seen to eat a fledgling Blue-black Grosbeak in Panama, and a Montezuma Oropendola (Psarocolius montezuma) was seen to kill a young Black-faced Grosbeak in Costa Rica.

Only an estimated 15% of 121 Northern Cardinal nestsmonitored in south-western Ohio were ultimately successful. The others were all known or suspected to have been preyed upon. Choice of nest-site appeared to have no effect on the final outcome. Instead, the birds' chief response to a large and varied guild of nest predators, including snakes, squirrels and crows, each with its own micro-habitat preference and search strategy, appears to be the ability to renest rapidly after failure. Cardinals may start a new clutch as little as four days after the loss of an earlier one.

Of 45Painted Bunting nests found by Parmelee in Oklahoma, only 15 produced fledglings, yielding a total of 37 young. In Kansas, Zimmerman found that nest predation was higher among Dickcissels nesting in old fields than for those in prairie, presumably because predators, especially snakes, are more common in old-field habitats. Predation reached a peak in July, the middle of the nesting season, when nest density was also at its height. Dickcissels that fail in a first nesting attempt may emigrate elsewhere. For example, 62% of Dickcissels studied in south-eastern Illinois relocated more than 10 km away before renesting, in one case moving more than 800 m. Renesting efforts began 4-15 days after failure, with an average interval of 8·5 days. Although successful Dickcissels generally did not renest, only 2% of those in the study abandoned nesting attempts after a first failure.

The period immediately after fledging, before the young have acquired flight skills, is the most critical for the survival of young. Rose-breasted Grosbeaks are incapable of flight for the first 2-5 days after leaving the nest, and move only short distances from the nest-site. L. C. Moore and his colleagues, in a study of a population of Rose-breasted Grosbeaks in fragmented forests in southern Ontario, found that 86% of total fledgling mortality occurred during the first week after fledging, mostly during the first five days. During this time the young usually remained perfectly still, even when approached closely. They seemed to rely for protection mostly on camouflage and parental defence. Older fledglings, when approached, reacted by moving higher in the vegetation, and tended to stay higher in the canopy, where predators may be less frequent. As the post-fledging period progresses and the young fly better and beg less, their risk of succumbing to a predator decreases and their likelihood of survival improves. Overall fledgling survival in this population was 0·62 for the three-week post-fledging period, a figure comparable to that for other small passerines.

Mortality rates for fledgling Dickcissels in the Missouri Valley were also highest during the first week after departure from the nest. Only 33% survived the first four weeks after fledging. Predation was a major cause of mortality, snakes, including garter snakes and fox snakes (Elaphe vulpina), being key predators of both nestlings and fledged young. Most mortality occurred when young birds were within 10 m of the nest-site, and it could be that the density of vegetation cover in the area near the nest is key to fledgling survival. In another studied population in south-eastern Missouri, only 56% of fledglings survived their first two weeks.

Fragmented landscapes frequently expose birds to higher rates of nest predation, as well as brood parasitism (see General Habits). L. A. Smith and her colleagues found that Rose-breasted Grosbeaks in heavily cut woodlots in southern Ontario tended to select lower, better-concealed nest-sites than did birds in lightly logged stands with a less open canopy. Such sites may have exposed the grosbeaks to higher levels of nest predation, and did expose them to higher rates of parasitism.

D. Morris measured the cost to female Indigo Buntings in fragmented forests in Missouri of having to replace a brood lost to a predator. She found that, although renesting could replace lost young, it did so at the expense of females' physical and physiological state. In the fragmented landscape, only 34% of females were able successfully to raise young on their first attempt, compared with 61% in contiguous forest. Birds that lost nests to predators often began another attempt 3-5 days later, perhaps before the females had had time to recover physiologically. Renesting buntings suffered poor body condition and high acute corticosterone levels, and produced young that were themselves in poor condition.

As mentioned earlier (see Systematics), the genera Parkerthraustes andSaltator are thought not to belong with the true cardinalids, and they are, therefore, treated separately in this section. It appears, however, that nothing is known about the nesting habits of the genus Parkerthraustes, the Yellow-shouldered Grosbeak, and, aside from Skutch's observations on the Buff-throated and Streaked Saltators in Costa Rica, the breeding biology ofSaltator has not been studied in detail.

The breeding season of saltators varies with species and geographical location. In central and northern South America, it may extend from March or April to August, although records outside this period exist. In Argentina and Uruguay, the season is primarily from October to January. The Greyish Saltator breeds between March and October in Trinidad, but between October and January in Argentina.

So far as is known, all saltators build an open cup-nest, although the nests of three of the species, theBlack-throated Grosbeak and the Black-cowled and Masked Saltators, remain to be described. Nests are constructed variously from twigs, leaves, roots and dry grasses, and lined with finer material. The nests of the Buff-throated and Black-headed Saltators have been described as bulky, and Skutch reported that the Buff-throated Saltator's nest was considerably bulkier than that of the Streaked Saltator.

Saltator nests are normally situated low in a bush or tree, in general not above 4 m, but occasionally they are higher, particularly in the case of the Lesser Antillean and Streaked Saltators. Of 22 Streaked Saltator nests studied by Skutch, 16 were placed between 1·2 m and 2·4 m above the ground, and one was as high as 6 m up in a mango tree (Mangifera). Buff-throated Saltators may nest in dense stands of bracken or compact tufts of tall grass in a pasture, and Orinoco Saltators will nest in cane. A Black-winged Saltator nest in north-western Ecuador was supported by pasture grass.

Female Buff-throated Saltators apparently build their nests alone. The male does, however, stay in attendance, bringing food to the female and singing responsively with her during building. Nest-building can take 3-6 days, but nests constructed later in the season are usually built in only three or four days.

The usual clutch size of saltators is two eggs or, less commonly, three. Egg-laying by Buff-throated Saltators can begin on the day after the completion of the nest or up to four days later. Once laying commences, eggs are laid on consecutive days. The first egg is normally laid in the hour following sunrise, the second one following some 26 hours later. Saltator eggs are blue, bluish-white or bluish-green, generally marked with blackish to brownish spots or streaks. In the case of both Buff-throated and Streaked Saltators, incubation is by the female alone, and is not continuous. Skutch recorded one female Buff-throated Saltator that, during a 12-hour observation period, devoted 66% of her time to sitting on the eggs, which she did in 18 bouts ranging from 6 to 52 minutes in length. Other females spent about the same amount of time in incubating, from 60% to 66%. Streaked Saltators incubate for similar periods, with longer sessions on rainy afternoons. The male Buff-throated Saltator may visit his mate during incubation, while the male Streaked Saltator is more likely to approach the nest when the female is absent. Both species, however, will feed the female at times as she incubates.

The eggs hatch after 13 or 14 days. Newly hatched Buff-throated Saltators have pink skin with a few tufts of long, olive-green down and a red mouth-lining. The young are fully fledged by about 12 days of age, and leave the nest at 13-15 days. For the first few weeks out of the nest, they remain in dense cover. A Streaked Saltator chick was fully fledged by 11 days and left the nest at 13 days.

Both parents, at least of the Buff-throated and Streaked Saltators, feed and attend the young, although apparently only the female broods. Streaked Saltators feed their chicks on a mixed diet of fruit and insects. Skutch observed a male Buff-throated Saltator that fed a nestling only two hours after the latter had hatched. Visiting parent Buff-throated Saltators mashed food items thoroughly in the bill before offering them to their nestlings. At two nests of this species, situated 2·4 m apart in coffee bushes, which Skutch watched over a total of 15 hours, the nestlings were fed 86 and 59 times and brooded for 193 and 226 minutes, respectively. Young Lesser Antillean Saltators are likewise attended by both parents. Apparent helpers at the nest have been recorded for the Black-throated Saltator (see General Habits).

Buff-throated Saltators in Costa Rica raise at least two broods in a season, and replace clutches lost to predation. Skutch recorded that nest predation was high for both Buff-throated and Streaked Saltators. In addition, Black-throated Grosbeaks and Green-winged and Greyish Saltators have been recorded as hosts of the Shiny Cowbird.



Cardinalidae includes not only resident species, but also long-distance, short-distance and altitudinal migrants. Mexican and Neotropical species are largely sedentary, although Crimson-collared Grosbeaks and Blue Buntings have been recorded well outside their normal range, and Rose-bellied and Orange-breasted Buntings are absent seasonally in some parts of Chiapas, in Mexico.

Glaucous-blue and UltramarineGrosbeaks are partial austral migrants. The Yellow Grosbeak is a summer visitor in the northern part of its range, and has wandered repeatedly to southern Arizona. The Black-backed Grosbeak may be a partial or altitudinal migrant in some parts of its range, but its movements are not well understood. On the Caribbean slope of Costa Rica, the Black-thighed Grosbeak is an altitudinal migrant, moving from higher-lying forest to lower elevations in numbers varying from year to year, probably in response to variations in food abundance.

Although Northern Cardinals and Pyrrhuloxias do not migrate, both species gather in flocks during the winter months, and introduced Northern Cardinals appear to have spread through the Hawaiian Islands at least partly under their own volition (see Relationship with Man). Black-headed Grosbeaks, Blue Grosbeaks and Varied and Painted Buntings are migratory over much of their range, but their breeding and wintering areas overlap, and southern populations may be sedentary. Southern populations of the Black-headed Grosbeak may shift altitudinally. In contrast, Rose-breasted Grosbeaks, Indigo and Lazuli Buntings and Dickcissels are complete migrants, leaving their northern breeding areas in autumn for wintering grounds in Middle America and, for Rose-breasted Grosbeaks and Dickcissels, South America. Indigo Buntings from Michigan may travel 2900-4000 km to reach their winter quarters.

Black-headed and Blue Grosbeaks, Lazuli Buntings and western populations of Painted Buntings are moult migrants (see Morphological Aspects). In the autumn, they largely transfer to staging areas in the south-western United States and north-western Mexico, where they moult before completing the journey southwards to the wintering grounds.

Migratory species breeding in eastern North America, namely the Rose-breasted Grosbeak, eastern populations of the Blue Grosbeak and Indigo Bunting, and Painted Buntings from the Mississippi Valley and Gulf coast, cross the Gulf of Mexico, although some Rose-breasted Grosbeaks fly around the gulf through north-eastern Mexico. Indigo Buntings flying northwards over the Gulf of Mexico in spring stop to feed along the southern coast of the United States before continuing on to their breeding areas.

Western species andwestern US populations of more widespread cardinalids migrate over land, though Lazuli Buntings appear on offshore islands in California during passage. Western migrants may still have to cross substantial barriers. As an example, Black-headed Grosbeaks crossing the Mohave Desert in autumn must cross 800-1100 km of desert terrain.

Lazuli and Indigo Buntings migrate along a broad front, rather than following mountain ridges. Indigo Buntings apparently fly downwind duringthe spring migration, both while crossing the Gulf of Mexico and while flying over land, and wind direction may influence the route that they follow.

Rose-breasted Grosbeaks and Indigo Buntings have broad breeding ranges in North America, and the extent to which eastern and western breeders mix on the wintering grounds is not entirely known. Ringing records suggest that eastern Rose-breasted Grosbeaks winter in Panama and northern South America, while more westerly ones spend the non-breeding season in Mexico and western Central America. Individuals from Ontario and New England have been recovered in western Central America, however, so it is evident that some mixing of populations does occur. The situation is similar for the Indigo Bunting, eastern and mid-western populations of which apparently follow parallel migration tracks and winter in separate areas. Indigo Buntings have been recorded as wintering in good numbers in Jamaica since the 1950s, this possibly representing a recent invasion of the island, as earlier observers failed to record them. One individual recovered in Jamaica had been ringed in Pennsylvania, in the east United States.

Eastern populations of the Painted Bunting migrate mostly to Florida and to a limited extent to the Gulf coast, the Bahamas and Cuba. More westerly breeders winter in Mexico and western Central America. Eastern Painted Bunting apparently show strong site-tenacity on their wintering grounds, the same individuals being trapped year after year in the same localities.

So far as is known, cardinalids generally migrate by night, although Black-headed Grosbeaks have been recorded as moving by day in the Mohave Desert, and Lazuli Buntings may make short-distance movements during the day after arriving on the breeding grounds. Indigo Buntings depart on migration flight at night, but sometimes continue to fly into the daylight hours. In mid-April, ffrench observed groups of up to a hundred Dickcissels as they departed after dark, one after another, from their roosts in Trinidad. Departures continued for several nights until the roosts were empty. At other times during the wintering season, the birds did not leave their roosts at night after settling unless they were threatened by fire.

While migrating, cardinalids travel singly or in mixed-species or single-species flocks. Flocks of more than 50 migrating Rose-breasted Grosbeaks have been recorded in Panama, and J. Van Tyne recorded Indigo Buntings as gathering in flocks in Guatemala during the weeks before they left to cross the Gulf of Mexico.

Spring migration is generally less protracted than that in autumn, probably because there is an advantage in securing territories and mates for adults that arrive early on the breeding grounds. There may, nonetheless, be considerable variation in the timing of spring migration among Indigo Buntings. Rose-breasted Grosbeaks have been estimated to migrate at an average rate of 38 km per day in spring and 80 km per day in autumn, including stopovers. The travel speed of migrating Indigo Buntings has been estimated at about 32 km per hour. Spring "overshoots", in which individuals occur far north of their normal breeding range in early spring, have been recorded for Blue Grosbeaks and Painted Buntings.

As is typical for North American passerines, males of migratory species tend to arrive on their breeding grounds ahead of the females, and older males before younger ones. The difference may be a matter of only a few days. Mean passage dates of Rose-breasted Grosbeaks captured on spring migration at Prince Edward Point, in Ontario, were 15th May for adult males and 18th May for yearling males and females. More brightly coloured yearling males tended to arrive earlier than dull-coloured birds.

Adult male Black-headed Grosbeaks arrive on their breeding grounds in New Mexico within 4-5 days of one another, while females start to arrive 5-7 days after the arrival of the first males. Yearling males reach the area some time later, 12-16 days after the first adult males and well after all adult males have settled, thereby avoiding the adults' peak period of aggressiveness. Unlike the Ontario Rose-breasted Grosbeaks, there is no apparent tendency for yearling males with more adult-like plumage to arrive earlier.

Lazuli Bunting males begin arriving in Utah, in the western United States, about four days before females and older birds two or three days before yearlings, although there is considerable overlap. There can be a gap of up to two weeks between the arrival of the first male Indigo Bunting and that of the first female, this apparently resulting from differences in pre-migratory fattening schedules and in times of leaving the wintering area. The average departure date recorded for males of all age-classes wintering in Jamaica was 16th April and that for females was 23rd April, a difference of a week. Arrival dates of Painted Buntings are less precisely known, but males were present in South Carolina one week before females and, in Georgia, adult males arrived 1-2 weeks before the arrival of yearlings.

The timing of autumn departure may also be staggered according to age and sex. Adult male Black-headed Grosbeaks leave the breeding grounds in late July or early August, but females and young of the year remain for several weeks longer, usually departing in mid-August. Adult male Lazuli Buntings leave their breeding grounds slightly before the females, and older birds slightly before younger ones, with juveniles of the year the last to depart. Adult males from the eastern population of Painted Buntings arrive on the wintering grounds in Florida before the females.

Prior to autumn migration, most Indigo Buntings increase their mass by some 50% as they accumulate subdermal and abdominal fat deposits. Those that do not do so may stop to feed en route; in order to cross the Gulf of Mexico, a bunting will need to expend about 30% of its body mass. Indigo Buntings wintering in Jamaica take on fat prior to the northward return migration, but those in the southern part of the wintering range do not; presumably, the latter stop to build up their fat reserves in Mexico before re-crossing the Gulf. As the birds move northwards towards their breeding grounds, they stop to feed, but they arrive leaner than they become later in the season.

Dickcissels gain extra fat prior toautumn migration both by eating more and by increasing their digestive efficiency so that fewer nutrients are lost in their excreta. In late March and April, the birds become hyperphagic on their wintering grounds, "overeating" in order to gain fat reserves. Metabolic adjustments allow them to put on fat more rapidly than do those individuals whose gain in mass is due to food intake alone. Dickcissels wintering in Trinidad may increase their weight by 56%. These behavioural and physiological shifts appear to be controlled by internal mechanisms, and will continue for at least a year even if the birds are kept in controlled environments with constant daylength and temperature.

In the 1960s, Emlen carried out studies of migratory orientation by Indigo Buntings and a single female Rose-breasted Grosbeak that exhibited migratory restlessness ("Zugunruhe"). Both species were able to orient themselves in the appropriate direction when exposed to the night sky. In experiments in a planetarium with captive Indigo Buntings, the birds oriented themselves in the appropriate direction as indicated by the sky projection, but they were unable to orient if no projection was shown. These results supported the hypothesis that the birds steer primarily by the patterns of stars in the night sky. Further experiments revealed that the buntings were able to adjust for stellar rotation as the nights progressed, suggesting that they have an internal clock that allows them to compensate for shifts in time.

Other mechanisms, however, must also be involved in migratory orientation. Emlen and his colleagues demonstrated in further experiments that, even when the night sky was visible, alteration of the magnetic field caused captive Indigo Buntings to shift direction. P. D. Sniegowski and his colleagues captured adults of this species in summer, and in the following spring released them some 1000 km from the capture site. The birds were able to return to their former breeding sites and reclaim their territories, suggesting that experience with familiar landscape cues may be important in informing them that they have arrived at their destination.

A few species, including in particular the Rose-breasted Grosbeak, have strayed to Europe, and the Indigo Bunting has been recorded in Russia. The likelihood that other members of the family will turn up as vagrants in places so far distant from the New World, however, is thought to be low.

Finally, with regard to the two probably non-cardinalid generaParkerthraustes andSaltator, neither the YelIow-shouldered Grosbeak nor any species ofSaltator is migratory, although some saltators are known to wander. In Ecuador, the Masked Saltator undertakes non-seasonal movements, perhaps in response to changes in the availability of cones of Podocarpus oleifolius (see Food and Feeding).


Relationship with Man

Cardinalids have attracted human attention, and admiration, for centuries. This is reflected in some of their local names. For example, in South America, the Black-backed Grosbeak is known asrey del bosque, meaning "king of the woods", whileCyanocompsa andCyanoloxia grosbeaks are known asreinamora, or "gypsy queen", the Ultramarine Grosbeak beingreinamora grande while the Glaucous-blue Grosbeak isreinamora chica.

Northern Cardinals have become human commensals in many areas. They are common and popular year-round visitors to suburban feeding stations, where, as M. M. Nice put it in 1927, "they rejoice us with their flaming beauty and their splendid song". At the other extreme, Black-backed Grosbeaks are reportedly killed by some fruit-growers, who regard them as a pest.

The Dickcissel has been hunted for food and heavily persecuted as a crop pest in Venezuela, where it is known asel pájaro arrocero, "the rice bird" (see Food and Feeding, Movement, and Status and Conservation). Hunters shoot these birds with guns and slingshots, enter their roosts at night and club the birds with sticks and bats, and even drive vehicles rapidly through flocks flying low across farm roads. In 1963, in Mexico, an unusual concentration of millions of Dickcissels near the city of Culiacán, in Sinaloa, wrought havoc on local wheat crops, and the farmers responded with firearms, explosives, slingshots, sticks and huge quantities of pesticides.

Cardinalids were well known to native Americans. A Yanomamo tale mentions that Yellow-green Grosbeaks eat hayi fruits. In the late sixteenth century, the artist and colonist John White recorded, on a watercolour, the Algonquian nameMeesquouns for the Northern Cardinal in North Carolina.

Cherokees invoked the Indigo Bunting to control the wind, possibly because of its blue colour, as they called upon the Eastern Bluebird for the same purpose. Cherokee men invoked the Northern Cardinal, the "Redbird", as part of a spell to attract women, as transcribed some time after 1890:


I am dressed as well as the Redbird

I am as handsome as the Redbird

I am as masculine as the Redbird

I can do as much as the Redbird

I can say as much as the Redbird

Dho:tsu! Dho:tsu:hwi!


The Cherokees believed that the Northern Cardinal was the daughter of the sun. According to a Cherokee legend, the cardinal was originally a brown bird, but, after he had helped a wolf to clear mud from his eyes-mud placed there by a trickster raccoon-the wolf rewarded him by painting his feathers red with pigment from a magical rock. In a variant of the story, the bird jumped into a magical pool, being careful not to submerge his mouth, which is why there is no red around the cardinal's beak. By the time his mate arrived, there was enough red paint left only to daub her in a few places, leaving her barely touched with red to this day.

Cardinals have been regarded as omens. Some southern native Americans believed that the sudden flight of a cardinal foretold the arrival of a guest. There is also a superstitious belief that a cardinal flying up is a sign of good luck, but one flying towards the ground is a portent of ill fortune.

The Northern Cardinal has been chosen as the state bird of seven American states, starting with Kentucky, in 1926, and followed by Illinois, after a poll of the state's schoolchildren, and Ohio, North Carolina, West Virginia, Virginia and Indiana. The bird has given its name to a number of sports teams, including the St Louis Cardinals baseball team.

Northern Cardinals have been kept in Europe since the sixteenth century. An account entitled "De Coccothrauste Indica cristata", accompanied by a woodcut apparently drawn from life, was published in 1600 inTomus Alter of theOrnitologiae of U. Aldrovandi (1522-1605). Even then the bird was known as "cardinalitus", presumably, as Aldrovandi notes, because of its red colour, the hue associated with ecclesiastical bishops. Aldrovandi passes on a report that a living specimen ate almonds and chickweed, became agitated on seeing its reflection in a mirror, plunged itself into water frequently, and imitated the voices of other birds, particularly the Common Nightingale (Luscinia megarhynchos). In 1731, M. Catesby noted that the bird was known in England as the "Virginia Nightingale", and was popular in Europe for its "beauty and agreeable singing".

The Painted Bunting hasfor long been known in aviculture as theNonpareil, a word meaning "without equal". According to Audubon, even in the early nineteenth century large numbers were trapped in Louisiana for the cagebird trade. People used a mounted male as a lure to catch the birds; a passing male Painted Bunting "dives towards the stuffed bird, with all the anger its little breast can contain. It alights on the edges of the trap for a moment, and throwing its body against the stuffed bird, brings down the trap, and is made prisoner." The birds were bought by middlemen for sixpence in New Orleans, but in England they fetched three guineas each in London. Few vessels left New Orleans in the summer months without carrying some Painted Buntings. Painted Buntings continue to be extensively trapped on their wintering grounds (see Status and Conservation).

All except one of thePasserina buntings are kept and bred in Europe. The single exception is the Rose-bellied Bunting, which apparently has never been imported into Europe, a rather surprising fact when one considers that the Orange-breasted Bunting, even though it has proven a difficult species to breed, is well known in trade.

Other members of the family commonly kept as cagebirds include the Ultramarine and Glaucous-blue Grosbeaks, the Vermilion Cardinal and the Green-winged Saltator (see Status and Conservation).


Status and Conservation

None of thespecies presently included in the family Cardinalidae is currently considered to be globally threatened, but four are listed as Near-threatened. These are the Painted Bunting, the Rose-bellied Bunting, the Masked Saltator and the Rufous-bellied Saltator. Although not classified as globally threatened, the Black-throated Saltator, acerrado endemic, is listed as "Vulnerable" in the state of São Paulo, in Brazil. In the 2007 WatchList published by the National Audubon Society and the American Bird Conservancy, Varied and Painted Buntings are listed in the "Yellow category", indicating that they are declining or rare.

The Masked Saltator is known from only a few disjunct localities in the central Andes of Colombia, southern Ecuador and north-central Peru, where it is rare and local (see Habitat, and Food and Feeding). L. M. Renjifo has suggested that its low population density in Colombia may be the result of its strong preference for highly localized patches ofPodocarpus oleifolius, a slow-growing primary-forest tree that has been overcut for its valuable wood. The species' known localities in Colombia and Ecuador are in heavily deforested areas, and in general the montane forests of the north Andes are under intense threat from conversion to agriculture and cattle pasture, mining and logging.

In addition to its global Near-threatened status, the Masked Saltator has been listed as "Vulnerable" in Colombia. Its most viable populations may be in Podocarpus National Park, in southern Ecuador, and in the Alto Quindí­o Acaime Natural Reserve-Los Nevados National Park-Ucumari Regional Park complex, in Colombia. Deforestation in the south-eastern buffer zone of Podocarpus National Park is expected to eliminate half of the remainingPodocarpus in the years to come.

The Rufous-bellied Saltator of the Andes of Bolivia and north-west Argentina is thought to have a relatively small, fragmented population containing in excess of 10,000 individuals, and to be in decline owing to loss of its temperate shrubland habitat. It is very local outside the Bolivian province of Cochabamba, where it is common in some areas, but it is common in one area in La Paz, and it has recently been found at several new sites. Native shrub species may be essential to its survival, but shrubland patches, includingPolylepis and alder (Alnus) stands, in its range are being fragmented and converted to agricultural and commercial forest use.

The little-known Rose-bellied Bunting isconsidered Near-threatened on the basis of its restricted Mexican range. Habitat degradation, including forest fragmentation, has been occurring within its area of distribution, but the effect of this on the species is unclear. In Mexico, this bunting is classified as "Threatened" in Mexico.

With a seemingly healthy global population, estimated at 3,600,000 in 2004, the Painted Bunting is still common in some areas. Nonetheless, there is evidence that it declined rapidly throughout its range from the mid-1960s until at least 2004-2005. North American Breeding Bird Survey data showed a significant decline of 2·5% annually between 1966 and 2002 in the continental United States (including five years of data from north-eastern Mexico), amounting to a total loss of 55% of the original population. National Audubon Society Christmas Count data from 1960 to 2003 also demonstrated a population decline. This decline has been blamed on habitat degradation and loss, cowbird parasitism in the case of the eastern population, and heavy trapping on the wintering grounds for the pet trade. Development of coastal swamp-thickets and woodland edges, loss of stands of shrubby wax myrtle (Myrica) and buckthorn (Rhamnus) in open-canopied pines and hardwoods, and replacement of native grasses, the seeds of which are an important food source, with sod grass have significantly reduced available habitat in the limited coastal range of the eastern population. Loss of riparian habitat in the south-western United States and in north-western Mexico affects the western population's mid-migratory staging and moulting areas. Brown-headed Cowbirds have only recently expanded into the breeding range of the eastern population, where parasitism rates now reach as high as 80%.

Although protected by law in the United States, Painted Buntings have been trapped legally in Mexico for at least the last 50 years. One probably conservative estimate suggests that, between 1979 and 2002, 5800 individuals each year-a minimum total of 100,000 overall-were legally trapped in Mexico for the domestic pet trade alone. In the 1984-1985 harvest season, trappers for the domestic trade were authorized to harvest 23,000 birds in five Mexican states. Painted Buntings have also been reported in local markets in Cuba, Costa Rica, Nicaragua and Guatemala, and illegal trade in the species has been reported in both Mexico and Florida.

In addition to capture for domestic markets, Painted Buntings have been trapped for export to markets in Europe, Asia and South America. The United Kingdom, Belgium, the Netherlands, Greece, Italy, Spain, Malaysia, Japan, Argentina and Paraguay have been recorded as major importers of the species. Painted Buntings reportedly fetch US$70 per pair on Asian and European markets. Except during the years 1982-1999, when it banned wildlife exports, Mexico has exported large numbers of the birds. Export figures reached at least 15,000 individuals during the 1974-1975 harvest season, and an estimated 12,000 during 2001-2002. In 2004, Mexico and the United States proposed that the Painted Bunting be included on Appendix II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), but the proposal was not accepted.

Several other cardinalids are trapped in varying numbers.Ultramarine Grosbeaks are common targets of cagebird-hunters in Brazil, where there is extensive illegal domestic traffic in wild birds. They were among the species most frequently confiscated from bird-traffickers in Rio Grande do Sul, in south-east Brazil, between 1998 and 2000.In Catolé do Rocha, a city in Paraí­ba, in thecaatinga region of Brazil where native bird-keeping is particularly common,a 2006-2007 survey of cagebirds in markets and private homesfound Ultramarine Grosbeaks to be the sixth-commonest out of 38 species recorded. During 2006, Green-winged Saltators and Ultramarine Grosbeaks were the third and sixth most frequently confiscated bird species, with 2730 and 916 individuals, respectively, in the Brazilian state of São Paulo. Glaucous-blue Grosbeaks have also been popular targets, as have some other saltators. Numbers of Golden-billed Saltators in Córdoba, in Argentina, are reportedly dwindling owing to the capture of adults for pets, as well as high nest predation in rural areas.

In northern Venezuela, the Vermilion Cardinal is a target for poachers. Rodrí­guez-Ferraro found that populations at three of six arid-zone sites studied in 2004-2005 had been reduced as a result of illegal trapping for the cagebird trade. The species' long-term survival may depend on the protection of sites where poaching is not yet a problem.

Although the Varied Bunting is apparently expanding its range in Arizona, and its global population has been estimated at 500,000-5,000,000 individuals, it has become a rarity in southern Texas and is listed as "Threatened" in New Mexico, where it has only recently established itself, breeding having been first recorded in 1972. This species' habitat may be threatened by development along the US-Mexican border and over much of its Mexican range.

Currently classified as of Least Concern by IUCN, the Dickcissel has been the subject of considerable conservation attention in the past. One observer even predicted in 1979 that the species would be extinct by the year 2000. In some part this concern has been the result of a failure to appreciate the species' opportunistic colonization and breeding strategies. Dickcissels are erratic and invasive, and they can appear for a short time in areas not suitable for long-term colonization. Those that invaded the south Canadian state of Manitoba in 1973, for example, nested in alfalfa fields but, when the fields were cut, any nests with young were presumably destroyed.

During the first half of the nineteenth century theDickcissel was common along the eastern seaboard of the United States from Cape Cod south to the District of Columbia, but by 1904 it had completely disappeared as a breeding species east of the Allegheny Mountains, in the west-central Appalachians. This disappearance has never been satisfactorily explained, although it is possible that the eastern population represented a recent colonization, sustained by the spread of hay crops in the region. It was suggested at the time that the increased use of mowing machines during the nesting period might have been responsible, but S. N. Rhoads had pointed out in 1903 that these machines were used also in the western part of the species' range. Rhoads suggested that the birds had not disappeared at all, but that the eastern breeders had simply shifted to the west, possibly as a result either of some event on the wintering grounds or of meteorological changes along the spring migration route.

Whatever thetrue explanation, Dickcissels began to reappear in the east in the 1920s. In 1928 they staged an extensive eastern invasion, turning up in the area from New Jersey south to South Carolina, but the species failed to re-establish itself as more than an occasional straggler and sporadic nester.

Whether these shifts were accompanied by a genuine change in numbers remains unclear. By the mid-1960s, however, Dickcissel populations began to decline in earnest. The North American Breeding Bird Survey recorded a 40% continent-wide drop in numbers between 1966 and 1996, although the population appears to have stabilized to some extent by the mid-1990s. Data from this ongoing survey indicate that the overall decline has slowed to insignificant levels in recent years, and in 2007 the species was removed from the National Audubon Society's Watchlist for United States Birds.

Nevertheless, decline is still continuing at regional levels. Dickcissel populations in restored grasslands in Missouri and Iowa are not self-sustaining. R. J. Fletcher and his colleagues found rates of decline in Iowa to be approximately 40% per year, nest predation being the largest contributing factor (see Breeding). Similarly, in a study published in 2008, K. With and her colleagues found that Dickcissel populations in theFlint Hills of Kansas and Oklahoma, an area that preserves the largest intact tall-grass prairie (about 2 million ha) left in the world, are continuing to decline at an estimated annual rate of either 11-21% or 19-29%, depending on the analysis.

Dickcissels adapt better than do some other grassland species, such as Henslow's Sparrow (Ammodramus henslowii), to changes in land use, although they are less productive than Henslow's Sparrows in prairie fragments in south-western Missouri. Studies of nesting success failed to locate evidence connecting their fall in numbers to conversion of original prairie habitat to agricultural land. The primary factor in the Dicjcissel's decline in the second half of the twentieth century appears to have been persecution by farmers on the wintering grounds in Venezuela, where the species is perceived to be a serious crop pest (see Food and Feeding).

S. D. Fretwell suggested that the decline may have been related to a skewed sex ratio, brought about because male Dickcissels, which are larger than females, were better able to feed on the large seeds of cultivated rice and sorghum. He predicted that the population would come to be dominated by males, and that this could lead to the collapse of the species. Basili and Temple found a sex ratio on the wintering grounds of 1·53 males to 1 female, much lower than Fretwell had predicted, and they were unable to establish either that this was due to female starvation or that it had changed significantly since the 1960s.

Instead, Basili and Fretwell blamed much of the decline on direct persecution. Most efforts by Venezuelan farmers to prevent crop damage in the weeks preceding the harvest involve noise-making, people either banging pots and pans in the fields or shooting air cannons or bottle-rockets. Some, however, use organophosphates such as parathion and azodrin to poison water-holes, spray feeding areas just prior to morning feeding bouts, and spray nocturnal roosts from the air or the ground. Spraying of nocturnal roosts in particular sometimes kills hundreds of thousands of birds in minutes. One farmer estimated that he had killed more than 1,000,000 birds on his property over the years.

Although the use of lethal control methods appears to have declined in recent years, Dickcissels are still vulnerable on thewintering grounds. A single roost, which could be destroyed in a few nights, can contain up to 3,000,000 million individuals, 30% of the species' entire global population. Alternative, non-lethal control methods, in addition to simply frightening the birds away, could include the use of repellent chemicals such as methiocarb and anthraquinone. Trials of these chemicals on captive Dickcissels reduced the birds' consumption of rice by between 70% and 90%. In 1998, the Venezuela Audubon Society held a workshop on the matter, and this resulted in the formation of a Ricebird Alliance, led by the society and the region's two most important farmers' groups, intended to find co-operative ways to address the problem.

In contrast to the efforts made to remove species from sensitive areas, there have been several attempts to introcuce cardinalids in places outside their natural ranges. European settlers introduced Northern Cardinals into Bermuda, probably from Virginia, perhaps as long ago as 1700, but certainly by the early nineteenth century. Thousands were reportedly trapped there for the cagebird trade during the latter century. J. M. Jones, who apparently thought that the birds arrived in Bermuda naturally, reported in 1859 that a "favourite sport of native boys" was to catch cardinals with traps made from lantana stems and baited with maize. The species is still common in Bermuda, despite a reduction in numbers since the 1950s following increased urbanization, the loss of native cedar (Juniperus bermudiana) forest and the introduction of House Sparrows (Passer domesticus), Common Starlings (Sturnus vulgaris) and Great Kiskadees (Pitangus sulphuratus).

Cardinals, including individuals from a mixture of races, have been repeatedly introduced into south-western California since 1880, but have failed to spread. They were introduced in Victoria, in Australia, without success in the 1860s or 1870s. Cardinals were released on Tahiti at some point after 1921 and were common there for a time, but the species had apparently disappeared by 1972. In the Hawaiian Islands, Hui Manu and other acclimatization societies released 300-350 Northern Cardinals on Kauai, Oahu and Hawaii between 1929 and 1931. As part of a "Buy-a-Bird" campaign during 1929-1930, schoolchildren in Hilo, on Hawaii, were encouraged to raise money to bring 163 cardinals to the island from Oahu. The birds were well established in the archipelago by the end of the 1930s, and subsequently spread, apparently unassisted, to Maui by 1943, Niihau by 1947, Molokai by 1951, Lanai by 1957, Kahoolawe before 1980 and, surprisingly, Nihoa, at least 275 km from the nearest source island, by 1983.

Attempts were made in the 1930s and 1940s to introducePasserina buntings in Hawaii. On Oahu, Indigo Buntings were introduced during 1934-1936 and Painted Buntings before 1937, by Hui Manu, and more than 425 pairs of Orange-breasted Buntings were introduced in 1941-1950; Orange-breasted Buntings were brought to Maui in the 1940s; and Indigo and Painted Buntings were introduced on Hawaii in May 1937. Although Orange-breasted Buntings did breed on Oahu in 1950, none of these introductions was successful. Grosbeaks, possibly Black-headed, were released, again unsuccessfully, at Pu'u Wa'a Wa'a Ranch, on Hawaii, in 1902.

Although habitat fragmentation can have a highly detrimental effect on nesting cardinalids (see Breeding), the members of this family often appear to adapt well to disturbed habitats and human presence. For example, Black-thighed Grosbeaks are common in human-dominated landscapes in Ecuador. Blue-black Grosbeaks more than doubled their numbers in burnt forest plots 10-15 months after an accidental understorey wildfire swept through the Tapajós-Arapiuns Extractive Reserve in central Brazilian Amazonia in 1997, although the same species disappeared fromterra firme forest fragments 1 ha in size, though not from larger ones, at the Biological Dynamics of Forest Fragments Project (BDFFP) near Manaus.

The effects of human activity on cardinalids may be more subtle. T. M. Jones and his colleagues have recently demonstrated that the plumage brightness of male Northern Cardinals in Ohio is lowered in areas of increasing urbanization, although the birds can compensate for this in more urban areas by feeding on the carotenoid-rich fruits of ornamental plants such as Amur honeysuckle (Lonicera maackii).

A number of species in North America have expanded their ranges following human alteration of the landscape. During the 1980s, after severe frosts killed thousands of acres of citrus groves in central Florida, the abandoned groves became overgrown with weedy vegetation. This converted them into ideal breeding habitat for Blue Grosbeaks and Indigo Buntings, and these species have substantially increased their breeding range in the state. Painted Buntings have undergone a similar but more limited increase.

The Northern Cardinal has achieved a considerable northward range expansion since the nineteenth century, the probable result of warming climate trends, increases in edge habitat created by human alteration of the landscape, and the popularity of winter feeding stations. The cardinals may have followed river valleys as they moved northwards, although they are not confined to riparian habitat in core areas of their range.

Northern Cardinals were first recorded in Ontario in 1849, in South Dakota in 1887, in Michigan in 1884 and in Minnesota in the late nineteenth century. They invaded Wisconsin in 1900 and again, more successfully, in 1910-1912, 1919-1920, 1927-1928 and 1932-1935, each invasion carrying them farther north in the state. They expanded "spectacularly" in New England during the 1950s. Cardinals expanded their range in Canada throughout the twentieth century, particularly after the late 1930s. They are now abundant residents in southern Ontario, having reached Ottawa, Sudbury and Thunder Bay by 1976. The species has also expanded its range in the south-west United States, spreading through the Colorado River Valley in Arizona and, marginally, south-eastern California.


General Bibliography

Akiko (1996), Aldrich (1948), Aldrovandi (1600), Aliabadianet al. (2009), Almeidaet al. (2006), Alsop (1979), Alveset al. (2010), Anderson, A.H. & Anderson (1946), Anderson, B.W. & Daugherty (1974), Anderson, J.F.et al. (1986), Anderson, M.E. & Conner (1985),Anon. (2003, 2004, 2007b, 2010b, 2010c, 2010e, 2010f, 2010g, 2010h, 2010i, 2010j, 2010k, 2010m), Antevs (1947), Audubon (1831), Averyet al. (2001), Baird (1964), Baker (1991, 1994, 1996), Baker & Baker (1988, 1990), Baker & Boylan (1995, 1999), Baker & Johnson (1998), Ballentine & Hill (2003), Bangs & Bradlee (1901), Barber & Martin (1997), Barlowet al. (2002), Barnett, Clarket al. (1998), Barreiraet al. (2007), Basili & Temple (1995, 1998, 1999a, 1999b), Bass (1979), Beddall (1963),Beecher (1953), Belcher & Thompson (1969), Beltzer (1988), Beltzer,Comini et al. (1999), Beltzer,Quiroga et al. (2004), Benediktet al. (2009), Berkeleyet al. (2007), Bernardi De León (2009), Besseret al. (1970), Blake, C.H. (1969, 1971), Blake, J.G. & Loiselle (1992), Bock (1964, 1994), Bodrati & Sierra (2008), Borges & Cardoso (1995), Borror (1961), Bosqueet al. (1999), Brackbill (1944, 1952, 1954, 1967), Bradley (1948), Brandtet al. (2009), Braun & Emanuel (1982), Breitwischet al. (1999), Brennan & Schnell (2005), Brower & Calvert (1985), Brown, B.T. (1994), Brown, H.C. (1920), Browning & Robel (1981), Burhans (1997), Burhanset al. (2000), Burns, K.J.et al. (2003), Burns, R.D. (1963), Butcheret al. (2007), Cáceres & Moura (2003), Callaghan (1983), Campet al. (2010), Campbell & Lack (1985), Cannellet al. (1983), Cant (1979), Cardiff & Remsen (1981), Carey (1982), Carey & Nolan (1979), Carling (2008), Carling & Brumfield (2008a, 2008b, 2009), Carlinget al. (2010), Carreraet al. (2008), Catesby (1731), Chapman (1927), Chesseret al. (2009), Christy (1942), Conneret al. (1986), Cracraft & Barker (2009), Crowell (1962), Crowell & Rothstein (1981), Dance (1986), Dawson (1948), Deane (1909), Dearborn (1996), Dearbornet al. (1998), Dechantet al. (2002), Demastes & Remsen (1994),Dickinson (2003), Dittus & Lemon (1969, 1970), Dixonet al. (2008), Dow (1969, 1970), Dow & Scott (1971), Downer (1972), Dunham (1965, 1966a, 1966b, 1966c), Eaton (2001), Echeverry-Galvis & Morales-Rozo (2007), Eckerle & Breitwisch (1997), Eguiarte & del Rio (1985), Eitniear (2006), Eitniear & Tapia (2000), Emlen, S.T. (1967a, 1967b, 1967c, 1971a, 1971b, 1972), Emlen, S.T.,Rising & Thompson (1975), Emlen, S.T.,Wiltschko et al. (1976), Estepet al. (2005), Faustino & Machado (2006), Fernández-Juricicet al. (2004), Ferreira & Glock (2004), ffrench (1967), Filliater & Breitwisch (1997), Filliateret al. (1994), Finck (1984), Fink & Brower (1981), Fisk (1974), Fjeldså & Rahbek (2006), Fletcher, N.H.et al. (2006), Fletcher, R.J.et al. (2006), Forbes (2010), Foster (2009), Francis & Cooke (1990), Fretwell (1973, 1977, 1979, 1986), Fretwell & Shane (1975), Friesen,Cadman & MacKay (1999), Friesen, Wyatt & Cadman (1999), Fuller (2010), Gabrielson (1915), Ganier (1941), Gardali & Nur (2006), Gardaliet al. (1998), Giocomoet al. (2008), Goller & Larsen (1997), Gomes de Almeida (2009), González-Ortegaet al. (2002), Goodpasture (1972), Gould (1961), Govoniet al. (2009), Grapputoet al. (2001), Greenberg & Marra (2005), Greene,Lyon et al. (2000), Greene,Muehter & Davison (1996), Greeney & Nunnery (2006), Groschupf & Thompson (1998), Gross (1921, 1956), Grubbet al. (1991), Halkin (1997), Halkin & Linville (1999), Hallinan (1924), Haney et al. (2001), Harmeson (1974), Hartupet al. (2000), Harvey (1903), Hatch (1983), Hawksley & McCormack (1951), Hayes & Sewlal (2004), Hellack (1976), Hellack & Schnell (1977), Hellmayr (1938), Henriqueset al. (2003), Hill (1987, 1988b, 1989, 1994), Hill & Kevin (2006), Hinds & Calder (1973), Holmes (1986), Hornet al. (2007), Howell (2010), Hudon (1991), Hugheset al. (1999), Humbird & Neudorf (2008), Hurley & Franks (1976), Hutto (1994), Hylton & Godard (2001), Ingold (1993),Íñigo-Elí­as et al. (2002), Ivor (1944), Jawor (2007), Jawor & Breitwisch (2003, 2004, 2006), Jawor & Gray (2003), Jawor & MacDougall-Shackleton (2008), Jawor,Gray et al. (2004), Jawor,Linville et al. (2003), Jenkins (1969), Jensen & Cully (2005), Jensen & Finck (2004), Johnston (1965, 1967, 1970), Johnston & Downer (1968), Jones, J.M. (1859), Jones, T.M.et al. (2010), Judd (1962), Kanegaeet al. (2008), Keyser & Hill (1999, 2000), Kilgo & Moorman (2009), Kimet al. (2008), Kirmse (1969), Kisslinget al. (2009), Klicka,Burns & Spellman (2007), Klicka, Fryet al. (2001), Klicka,Johnson & Lanyon (2000), Komar & Thurber (2003), Kopachena & Crist (2000a, 2000b), Krech (2009), Kroodsma (1974a, 1974b, 1975), Lanyon & Thompson (1984, 1986), Larsen & Goller (2002), Laskey (1944, 1969), Leberman (1984), Lemon (1965, 1966, 1967, 1968a, 1971, 1974), Lemon & Chatfield (1971, 1973), Lemon & Herzog (1969), Lemon & Scott (1966), Leopold & Enyon (1961), Leslie (2009), Leston & Rodewald (2006), Langley (1976), Lever (2005), Levey (1987), Levy (2009), Linville & Breitwisch (1997), Linvilleet al. (1998), Lloydet al. (1998), Locke (1965), Lockwood (1995), Logan (1951), Loiselle & Blake (1992), Long, C.A. (1963), Long, C.A.et al. (1965), Long, J.L. (1981), Lovette & Bermingham (2002), Lowther et al. (1999), Machado (1999), Maddox & Bollinger (2000), Maneyet al. (2008), Marchant (1960), Margoliashet al. (1991, 1994), Marler & Slabbekoorn (2004), Martin (1993), Martñ­nez-Meyeret al. (2004), McAtee (1908), McElroy & Ritchison (1996), McGraw,Hill & Parker (2003), McGraw,Hill, Stradi & Parker (2001), McNair (1990), McNair & Forster (1983), Meanley (1963), Mettler & Spellman (2009), Michener & Michener (1951), Moermond (1981, 1983), Moermond & Denslow (1985), Monson (1997), Mooreet al. (2010), Morris (2005), Morton (1975), Muehteret al. (1997), Mulvihill (1988), Munson & Robinson (1992), Nealen & Breitwisch (1997), Nice (1927), Nolan (1965), Nores & Cerana (1990), Olson & Alvarenga (2006), O'Neill & Schulenberg (1979), Ortega & Hill (2010), Ortega & Ortega (2003), Osborne (1992), Overmire (1962), Parker (1982), Parkes (1985), Parmelee (1959, 1964), Parrish (1997), Payne (1981, 1982, 1983, 1991, 1996, 2006), Payne & Payne (1989, 1990, 1993a, 1993b, 1996, 1998), Payne & Westneat (1988), Payne, Payne & Doehlert (1987, 1988), Payne,Thompson et al. (1981), Paynter (1970), Pearman (1997), Peer (2010), Perdue (1979), Phillips (1994), Piratelli & Pereira (2002), Pitzrick & Pitzrick (1992), Poulinet al. (1994), Prumet al. (2009), Pyle & Pyle (2009), Quay (1987, 1989), Ragusa-Netto (2001, 2002), Rahmiget al. (2009), Raikow (1978), Reese (1975), Remsen (1997b), Remsen & Parker (1984), Remsen,Cadena et al. (2010), Remsen,Schmitt et al. (1988), Renjifo (1991), Rhoads (1903), Rice, J.O. & Thompson (1968), Rice, O.O. (1969), Richet al. (2004), Richmondet al. (2007), Ricklefs (1976), Ridgely& Tudor (1989), Riedeet al. (2006), Rioset al. (2008), Risch & Robinson (2006), Ritchison (1981, 1983a, 1983b, 1983c, 1983d, 1985a, 1985b, 1986, 1988, 1997), Ritchisonet al. (1994), Riveraet al. (2008), Riverset al. (2010), Robbinset al. (1985), Rodrñ­guez-Ferraro (2008), Rodrñ­guez-Ferraro & Blake (2008), Rodrñ­guez-Ferraro & Trujlllo (2004), Rodrñ­guez-Ferraro et al. (2007), Rohwer (1986), Rohweret al. (1980), Rothstein (1973), Russell (1951), Rutter (1958), Salgadoet al. (2008), Samuelson (1980), Sandercocket al. (2008), Sarmiento (2010), Saueret al. (2008), Schartz & Zimmerman (1971), Schifter (1984), Schooket al. (2008), Schueller (2000), Scott (1998), Scott & Lemon (1996), Scottet al. (1987), Sealy (1976), Seutinet al. (1993), Sgueo (2009), Sharpe (1888), Shiovitz (1975), Shiovitz & Thompson (1970), Short (1975), Shumanet al. (1989, 1992),Sibley (1996), Sibley & Ahlquist (1990), Sibley & Monroe (1990, 1993), Sibley & Short (1959), Sick (1993), da Silva (1990), Skutch (1954, 1972, 1980a), Slud (1964), Smith, A. (1927), Smith, C.E. (1966), Smith, L.A.,Burke et al. (2006), Smith, L.A., Nolet al. (2007), Sniegowskiet al. (1988), Sorianoet al. (1999), Sprunt (1968), Stabler (1959), Stanback & Powell (2010), Steadman & McKitrick (1982), Stiles & Skutch (1989), Stoddard & Prum (2008), Stokes (1993), Stone (1927), Stoner (1947), Stotz (1992), Stoufferet al. (2009), Strecker (1893), Stutchbury & Morton (2001), Stutchbury et al. (2005), Sugiyama (2001), Sutton (1950, 1967), Sweeneyet al. (2004), Taber (1947), Tamplinet al. (1993), Tayloret al. (1989), Temple (2002), Thiollay (2002), Thomasset (1915), Thompson, C.F. & Lanyon (1979), Thompson, C.W. (1991a, 1991b), Thompson, C.W. & Leu (1995), Thompson, W.L. (1968, 1969, 1970, 1976), Tobias & Williams (1996), Todd (1923), Toledo & Hernández (1979), Tordoff (1954), Tubaro & Lijtmaer (2006), Tweit & Thompson (1999), Van Tyne (1932), Vondrasek (2006),Vu Halinh et al. (2007), Waide & Hailman (1977), Walk,Kershner et al. (2010), Walk,Wentworth et al. (2004), Wasser & Sherman (2010), Watadaet al. (1995), Weldon & Haddad (2005), Wellset al. (2008), West (1962), Westneat (1987a, 1987b, 1988b, 1989, 1990), Weston (1947, 1955), Whiteheadet al. (2000), Whittaker (1998), Wible (1974), Willoughby (1992), Winter (1999), Winteret al. (2000), Wiseman (1977), Withet al. (2008), Wolf (1971), Wolfenbarger (1999a, 1999b, 1999c), Wyatt & Francis (2002), Yamaguchi (1998a, 1998b, 2001),Yen Chunewei (1989), Young, B.E. (1991), Young, H.et al. (1941), Yuri & Mindell (2002), ZhangLuping et al. (2004), Zimmerman (1965a, 1965b, 1965c, 1966a, 1966b, 1971, 1982, 1983, 1984), Zimmerman & Finck (1989).