Italian Ornithological Web Site
by Alberto Masi since 1996

WHAT IS . . .

Phylogeny is to the ancestry of groups of organisms as genealogy is to the ancestry of human families.  The reconstruction of a human genealogy involves finding evidence of genetic relatedness from various sources, including family history and documents.  Phylogeny reconstruction must use fossils, similarities and differences in structure and other sources of information.  DNA comparisons became available as the biochemists who study nucleic acids, of which DNA is composed, developed methods for comparing the DNAs of different species representing different groups of organisms.  The following quotation is from the Introduction to "Phylogeny and Classification of Birds" (Sibley and Ahlquist 1990, p. 3).

"The plants and animals of Earth evolved from a single origin of life and they have had a single evolutionary history, or phylogeny.  The reconstruction of this phylogeny has been a challenge to biologists since Darwin (1859) recognized the principle of descent with modification.  The pattern of descent is genealogical; organisms are related to one another by the degree to which they share genetic information.
 The pattern of phylogeny is like that of a huge bushy tree with a short trunk that branched again and again, until the uppermost branches end in terminal twigs.  The height of the tree is relative to time, but the tree has a ragged top because different branches evolve at many different rates; the lengths of the branches reflect those rates.  Each branch node records a divergence event when one species split into two.  The horizontal spread of the branches represents the diversity of life, and the topmost twigs are the individual organisms living today.  The branches below the living twigs at the top are dead; they represent the ancestors, which, except for fossil remains, are invisible to us.  Most of the branches are broken off below the top; they are the extinct lineages.

The challenge is to reconstruct the branching pattern of the tree and, if possible, to date each branching event.  If we could do these things for all organisms we would have reconstructed the phylogeny of life on Earth.  Our present goal is to reconstruct the small cluster that represents the phylogeny of the approximately 10,000 living species of birds."
WHAT IS A SPECIES?:  Definitions and the "species problem".
This question remains controversial and, indeed, enigmatic.  We assign names to populations of organisms
and propose definitions of taxonomic categories based on various criteria, but the "species concept" is slippery because there are so many examples in nature of populations that refuse to fit our definitions.  Only partly as a joke, one definition is that a species is "whatever a competent taxonomist says is a species."  That only changes the debate to the definition of a "competent taxonomist" -- which usually means only the speaker!

 A widely-used definition is the "Biological Species Concept" which says that species are populations that
are interbreeding or potentially can interbreed, and which are reproductively isolated from other such groups by reproductive isolating mechanisms.  But there are numerous situations that defy the "potentially" aspect of this definition.  For example, how should we treat populations that are allopatric (not in physical contact), but which are so much alike that we think that they would interbreed if they came into contact (sympatric)?  For example, populations of morphologically similar sedentary birds isolated from one another on islands or mountain tops.  Again, opinion enters the definition because they might not interbreed and interbreeding in captivity does not prove that the two kinds would do so in the wild where mate choice is not limited.  If two species are sympatric and never interbreed, they are clearly different species, but how should we treat occasional hybrids between sympatric "species" which usually do not interbreed?  Hybrids present problems for every species concept.

Based on the Biological Species Concept, Ernst Mayr and Dean Amadon counted 8,590 species in 1951 (Amer. Mus. Novit. No. 1496) and 8,600 is often cited as "the number" of species of living birds.
Richard Bowdler Sharpe, curator of birds in the British Museum during the latter 1800's and early 1900's, did not accept the subspecies concept, thus what others were calling subspecies, Sharpe called species and used only binomials.  In his Handlist of the Genera and Species of Birds, vol. 5, p. xii, he listed 18,939 species.
A similar number might be attained by the recently-proposed "Phylogenetic Species Concept" that considers any population to be a species if the members are "diagnosably different" from those of other populations and that share a parental pattern of ancestry and descent.
With this definition the number of species of  birds-of-paradise rises from 43 to 90 and one estimate of the number of living species is over 18,000.

Another approach is the "Recognition Species Concept" that defines species as "groups of individuals that share a common fertilization system or specific-mate recognition system."  This concept emphasizes the factors that ensure interbreeding within a group of individuals, rather than the factors that reduce or eliminate breeding between members of different species, as in the biological species concept.
 Different definitions yield different numbers of species and the non-specialist comes to believe that avian systematists are merely engaging in a game, but there is serious science in all of this for we are trying to fit Nature into a system to improve our understanding of the processes and products of evolution and Nature is complex.  Some changes are arbitrary, but more often they are based on new information.  In many cases detailed field observations of behavior and vocalizations, and laboratory studies that compare proteins or DNAs, are clarifying the relationships among populations previously known only from museum specimens.  So, "splitting" and "lumping" will continue as we try to make Nature fit our concepts.  Of course, we like to think that we are making our concepts fit Nature.

Burt Monroe and I "recognized" 9,672 "species" in our 1990 book (Distribution and Taxonomy of Birds of the World) but in Monroe and Sibley (1993.  A World Checklist of Birds) we list 9,702 and this book-on-diskettes lists 9,730.  How can this be?  The answer is that we don't know how many "species" of birds there are because the definition of a species is so elastic and our ignorance of what happens in nature is so profound.  One or two new species are discovered in the wild each year, but the number of species "recognized" changes mainly because populations previously viewed as subspecies are raised to the rank of full species.  This will be apparent to anyone who uses the information in this Birds of the World.  In many accounts the phrase "May be a separate species" will be found.  In others:  "May be conspecific with ...." or a similar statement indicating another opinion.  Among the most "difficult" groups are the species of Scytalopus (Tyranni; Rhinocryptidae) and the numerous morphologically variable "subspecies" of the Golden Whistler (Pachycephala pectoralis: Corvidae; Pachycephalinae).

In 1984, when Burt and I began our collaboration on "Distribution and Taxonomy of Birds of the World", we decided to follow the definition of the "biological species concept" in deciding the boundaries of species, but to modify it to avoid some of the problems associated with decisions about "potential" reproductive isolation.
We decided to treat as species, populations that have been or can be considered to be "semispecies, allospecies, paraspecies, and most "incipient species"" -- translated into plain English this means that if proof of the ability to interbreed is lacking, or if the populations in question are geographically separated, they may be treated as allospecies, which are usually grouped into a "superspecies" containing the allospecies that we think represent a closely-related group.  Under the "biological species concept" many allospecies would be considered subspecies of the species that was the first to have been described and named.  The decision to use the allospecies concept (plus ca. 50-60 new species) is what increased our estimate of the number of species from the 8,590 of Mayr and Amadon (1951) to the estimate of 9,702 (Monroe and Sibley, 1993) and to 9,730 in the present classification.
Thus, most of the increase from 8,600 to 9,730 resulted from the transfer of ca. 1000 "subspecies" to the list of "species" -- mostly as allospecies.  However, during this time many "species" were converted ("lumped") into subspecies, so the number of changes is considerably more than 1000.

For example, consider the history of a "new subspecies" of vireo I described in 1940 (Condor 42:255-258), which I "discovered" in a tray of specimens in the Museum of Vertebrate Zoology at the University of California in Berkeley.  The specimens had been collected in the Victoria Mountains at the tip of Baja California by Chester C. Lamb, who collected for Joseph Grinnell.  The birds differed in color and size from their nearest relatives in California, but they were clearly "Warbling Vireos".  I named them Vireo gilvus victoriae -- but
the woods of the Victoria Mountains are isolated by the deserts of central and northern Baja California from Warbling Vireos in California -- so, has V. g. victoriae become a "good species" -- Vireo victoriae -- by our definition?  The answer seems to be that it is an "allospecies" -- a member of the Vireo gilvus "superspecies",
but geographically isolated from other members of the gilvus group.  Allospecies are often treated as species in this book, so Vireo victoriae is only one among many.  This raises the following question --.


WHAT ARE SUBSPECIES?:  Definition and discussion.

Subspecies are defined as "geographic subdivisions of species".  Subspecies are often called "geographic races" or just "races".  The definition is often elaborated with a statement that says or implies that such subdivisions either "intergrade" where they are in contact, or that it is assumed that they would if they came into contact -- so, we are back at the "potential" aspect of the biological species concept.  "Intergrade" implies a gradient of characters from one subspecies to the next, usually with a partial barrier causing the semi-isolation between them.  Patterns of this kind, often involving many subspecies, are quite common.  Subspecies usually differ slightly in plumage color and/or size, often associated with a climatic gradient or with barriers that inhibit gene flow and permit local differentiation.  So, subspecies are defined by geography and species by the ability to live in sympatry with close relatives and not interbreed in the wild -- but there are no absolute criteria to separate these human-imposed definitions.  We use them for convenience and "recognize" species, subspecies, allospecies, semispecies, etc. for lack of better ways to describe the complexities produced by evolutionary processes.  This complexity is part of the evidence for evolution by natural selection.  For examples of complex problems see the Sage Sparrow (Amphispiza belli) and the account of the studies by Ned Johnson, and the study of the Fox Sparrow (Passerella iliaca) by Robert Zink.  Subspecies are not the only problem taxa -- what about interspecific hybrids?

HYBRIDIZATION IN BIRDS:  Definition and examples

Interspecific hybridization in birds -- a subject to which I devoted a decade (1946-1956) and many days in the field, collecting specimens and observing the birds and their habitats in Mexico, the Great Plains and Colombia.  In Mexico I studied two species of towhees that live side by side in Oaxaca and seem never to hybridize.  On the western slope of Mount Orizaba in eastern Puebla they are in contact and occasional hybrids are found.  On Cerro Viejo near Lake Chapala, Jalisco,  they form a "hybrid swarm" in which all individuals show evidence of interbreeding and no two are exactly alike.  Across the mountainous backbone of central Mexico from near Mexico City to west of Lake Pátzcuaro there is a gradient in their characters from one end to the other.  These two species are the Rufous-sided Towhee (Pipilo erythrophthalmus) and the Collared Towhee (Pipilo ocai).  They are remarkably different in plumage pattern, color and size.  The Rufous-sided occurs from southern Canada to Guatemala; the Collared Towhee occurs only in the mountains of central and southern Mexico, usually in the undergrowth of pine-fir forest above the brushy oak-woodland habitat of the Rufous-sided, but their habitats overlap in many places.  In several areas the Collared Towhee occurs alone without any contact with the Rufous-sided.  Thus, they are both sympatric and allopatric; they interbreed in some places and not in others -- shall we call them two "species"-- "semispecies" or "subspecies" of one species?  I cut this Gordian Knot by calling them "species" with the midpoint of the gradient between Mexico City and Lake Pátzcuaro as the point of distinction -- to the west = ocai, to the east = erythrophthalmus.  However, this is not a solution, only a decision to make this intractable situation fit our concept of species.

There are many other places where two "species" (or "semispecies") of birds hybridize, often where one is at the edge of its distribution where conspecific mates are rare and a closely-related species is present.  Mallards hybridize with Pintails and other species of Anas and hybrids between the Capercaillie and the Black Grouse are so common in Scandinavia that hunters have given the hybrids a special name.  There are thousands of records of avian interspecific hybrids and it has been estimated that about 10% of the species of birds hybridize, at least occasionally, with close relatives.  Some, but not all, hybrids are sterile.  Only closely-related birds can, and do, interbreed, so hybridization tells us something about genetic relatedness and about how species evolve.  It also makes it difficult to use a species definition based on the ability to interbreed.


Latin Names:  The Latin names for animals are governed by an International Code of Zoological Nomenclature which is a system of rules and recommendations authorized by the International Congresses of Zoology.  The Code is adjudicated by the International Commission on Zoological Nomenclature whose members represent different groups of animals.  The members of the Commission meet at intervals to render opinions about the proper application of the Code to particular problems.  Thus, the Latin names for birds are not haphazardly chosen and applied, but are subject to a set of rules to reduce confusion.  The Code "is a set of criteria to be met in giving to an animal, or to a taxonomic group of animals, a scientific name, with its proper reference of author and date; and to regulate inter se names that have been given in the past."  This quotation is from the Introduction to the 1961 edition of the Code -- which continues:

"The assignment of a unique and distinct name in the modern meaning, by which to identify each kind of animal began in 1758.  It was in that year the Swedish naturalist Carolus Linnaeus (who was enobled in 1761 as Carl von Linné) brought out the 10th edition of his Systema Naturae.  In this he extended as a uniform procedure for animals a system others had used for restricted groups and he had earlier established for plants, namely, to give each species a simplified name consisting of only one word plus the generic name.  Thus, the dog became Canis familiaris, its generic name showing that it had certain readily identifiable characters in common with, for instance, those of the wolf, Canis lupus, and the jackal, Canis aureus."

The American Ornithologists' Union adopted a code of nomenclature for birds in 1885 and other organizations did so during the same period.  The International Code was formulated and adopted in the latter 1800's by a series of International Zoological Congresses and by 1905 the Rules were close to their present form.  The 1958 edition superseded earlier editions and the 1961 edition is a thin volume of 176 pages, including Recommendations, Index, etc.
 The Preamble states that:  "The object of the Code is to promote stability and universality in the scientific names of animals, and to ensure that each name is unique and distinct."

"Priority is the basic principle of zoological nomenclature."  This means that the first published name for any taxon (=a named unit) is the one that must be used, but there are some exceptions to this rule, such as preserving a long-accepted name even if an earlier one is discovered.  So, the Code is a bit elastic, but it serves to keep order in what had been chaos prior to ca. 1900.
 Linnaeus' 10th edition of his Systema Naturae was taken as the starting point for binomial nomenclature, at least in part because it was the first publication to use just two names for each species.  Names proposed before 1758 need not be considered, those published in the 10th edition and after 1758 must be considered.  The scientific name of an animal must be in Latin or Latinized, is composed of two units -- genus and species -- and must be written in italics -- Passer domesticus, Turdus migratorius, etc.  Thus, using italics for scientific names is not a whim, it is one of the Rules of Zoological Nomenclature.  There are Rules for the formation of group names -- Families, Orders, etc.  There are XVIII sections containing 87 Articles that occupy 91 pages in the 1961 edition.

The Latin names are relatively stable because of the application of the Rules.  The use of generic names provides an indication of probable close evolutionary relationship among the species placed in the same genus, and the higher categories are designed to accomplish the same thing for a set of nested groups -- Class, Order, Family, Subfamily, Genus, Species, Subspecies -- and often others.  Subspecies are given a third name, e.g., Turdus migratorius propinquus for the western subspecies of the American Robin.  The "typical" subspecies is the one that first received the species name, e.g., the eastern subspecies of the American Robin is Turdus migratorius migratorius -- usually written as Turdus m. migratorius and Turdus m. propinquus.



English Names:  There is no set of universally-applied rules to govern the use and application of English names, or for vernacular names in other languages.  English names often contain little or no information about relationships and may be misleading.  The lack of control results in many different names being used for some species and the same name being used for different species.  There are at least 67 different names for the Ruddy Duck used in North America and the Flicker is known by many local names.  Small birds with reddish underparts are called "robins" in English-speaking countries, worldwide, but many are not even distantly related to the English Robin (Erithacus rubecula).  Some progress is being made through committees appointed by the International Ornithological Congresses.  Committees for English, French and German were appointed in New Zealand in 1990 at the 20th I.O.C.


The Father of the Classification:
Hans Friedrich Gadow (1855-1928)

Hans Friedrich Gadow (1855-1928) may be characterized as the Father of the Wetmore Classification and Grandfather of the Peters Classification, because James L. Peters adopted Wetmore's 1930 classification when he began his Checklist of Birds of the World series in 1931.  Gadow was born in Germany and studied anatomy under Haeckel and Gegenbaur, famous names in the history of comparative anatomy.  Gadow published on the coiling patterns of the intestinal tract of birds and on many other aspects of avian anatomy.  He went to England to work with Richard Sharpe on the Catalogue of Birds, of which he compiled volumes 8 and 9 on passerines.  He became a British citizen, married the daughter of the Prof. of Physics at Cambridge, and spent the rest of his life as Curator of the Stricklandian Collections and Reader on the morphology of vertebrates at Cambridge University.  In 1892 Gadow published a new classification based in part on Maximilian Fürbringer's two massive volumes on avian anatomy (1888), plus other sources and his own research based on "about forty characters from various organic systems."  Gadow's forty characters, and the method he used to develop his classification, are important for an understanding of the classifications in use since 1893.  His "List of Characters employed in determination of the Affinities of various Groups of Birds" follows, with definitions and comments in parentheses.

A.  Development.
 Condition of young when hatched: whether nidifugous (precocial) or nidicolous (altricial); whether naked or downy, or whether passing through a downy stage.
B. Integument.
 Structure and distribution of the first downs, and where distributed.
 Structure and distribution of the downs in the adult: whether absent, or present on pterylae or on apteria or on both.
 Lateral cervical pterylosis: whether solid or with apteria.  (Pterylosis = feather distribution).
 Dorso-spinal pterylosis: whether solid or with apterium, and whether forked or not.
 Ventral pterylosis: extent of median apterium.

Aftershaft: whether present, rudimentary, or absent.  (Aftershaft = double feather).
 Number of primary remiges.  (Remiges = primaries and secondaries - flight feathers).
 Cubital or secondary remiges: whether quinto- or aquinto-cubital. (Refers to the presence or  absence of the fifth secondary, also called eutaxic = 5th present; diastataxic = 5th absent.
 Oil-gland: present or absent, nude or tufted.
 Rhamphotheca: the horny bill sheath covering the bony maxilla and mandible; whether simple  or compound, i.e., consisting of more than two pieces on the maxilla.

C. Skeleton.
 Palate: Schizo-desmognathous.  Nares, whether pervious or impervious, i.e., with or without a complete solid naso-ethmoidal septum.  (Thomas Henry Huxley (1867) defined four palatal types on the basis of the relationships among the vomer, pterygoids and palatine bones, viz., Dromaeognathous; Desmognathous; Schizognathous; Aegithognathous.  A fifth type, Saurognathous, proposed by W.K.  Parker, is found in some woodpeckers).
 Basipterygoid processes: whether present, rudimentary, or absent; and their position.
 Temporal fossa: whether deep or shallow.

Mandible: os angulare, whether truncated or produced; long and straight or recurved.
 Number of cervical vertebrae.
 Haemapophyses: the ventral processes or extensions of the cervical and thoracic vertebrae; occurrence and shape.
 Spina externa and spina interna sterni: = processes on the sternum or breastbone; occurrence, size, and shape.
 Posterior margin of the sternum: shape of.
 Position of the basal ends of the coracoids: = bones from sternum to shoulder in birds; whether separate, touching or overlapping at the basal end which is at the sternum.
 Procoracoid process: its size and the mode of its combination with acrocoracoid.

Furcula: the joined clavicles = "wishbone": shape; presence or absence of hypocleidium (process at base of furcula) and of interclavicular process.
 Groove on the humerus for the humero-coracoidal ligament: its occurrence and depth.
 Humerus: with or without ectepicondylar process.
 Tibia: the leg bone between femur and tarsals: with bony or only with ligamentous bridge, near its distal tibio-tarsal end, for the long extensor tendons of the toes; occurrence and position of an intercondylar tubercle in vicinity of the bridge.
 Hypotarsus: the structure of the head of the tarso-metatarsus, the long bone between the foot and the tibia: formation with reference to the tendons of the long toe-muscles whether (1) simple, if having only one broad groove; (2) complex, if grooved and perforated; (3) deeply grooved and to what extent, although not perforated.

Toes: number and position, and connexions (sic).
D. Muscles.
 Garrod's symbols of thigh muscles A B X Y,--used, however, in the negative sense.  (Refers to the presence or absence of certain muscles of the thigh region.  J.  George and A.  Berger made an extensive study of pelvic musculature in 1966 and concluded that pelvic muscle formulae alone are of limited value for the determination of phylogenetic relationships).
 Formation of the tendons of the m. flexor perforans digitorum (the tendons that are attached to the toe bones); the number of modifications is 8 (I.-VIII.) according to the numbering system in Bronn's Vögel, p. 195, and Fuerbringer, p. 1587.

E.  Syrinx.
 The sound-producing 'voice-box' at the junction of the bronchi.  Location as tracheal, broncho-tracheal, or bronchial.  Number and mode of insertion of syringeal muscles.
F.  Carotids.
 The main arteries from heart to head.  If both right and left present, typical: or whether only left present, and the range of modifications.  (Carotid patterns vary widely and were studied in detail by Fred Glenny in the 1950's -- his results indicate that numerous exceptions and special cases render adult carotid patterns useless as the basis for classification).
G.  Digestive organs.
 Convolutions of the intestinal canal.  Eight types, numbered I.--VIII., according to Bronn's Vögel, p. 708, and Proc.Zool.Soc. of London 1889, pp. 303-316.  (Later studies by Frank Beddard (1911) found little evidence of phylogenetic information; rather that the intestinal convolutions are related to food habits; their taxonomic value is nil.  Others have been unable to replicate some of Gadow's types because of poor definitions).

Caeca: = appendices of the intestinal tract - blind diverticulae like the human appendix; whether functional or not.
 Tongue: its shape.
 Food.--Two principal divisions, i.e., Phytophagous or Zoophagous, with occasional subdivisions such as Herbivorous, Frugivorous, Piscivorous, Insectivorous, etc.
List of Characters Employed occasionally.
Shape of bill.
Pattern of colour.
Number of rectrices; and mode of overlapping of wing-coverts, according to Goodchild (P.Z.S. 886, pp. 184-203).  (Rectrices = tail feathers).

Vomer.  (= bone in palate).
Pneumatic foramen of humerus.  (= opening from humeral head into hollow shaft).
Supraorbital glands.  (= salt-excreting glands located above the eye socket).
Penis.  Present in some groups, e.g., Ostrich, some ducks.
Certain wing-muscles according to Fuerbringer.
Mode of life: Aquatic, Terrestrial, Aerial, Diurnal, Nocturnal, Rapacious, etc.
Mode of nesting: breeding in holes.
Structure of eggs.
Geographical distribution.

 These were the characters used by Gadow to evaluate degrees of relationship among major avian groups.  He compared the conditions in each group with those in every other group and scored the results as the number of agreements vs disagreements -- the higher the agreement score, the more closely related the groups.  Sibley and Ahlquist (1990:205-224) examined six of the characters used by Gadow and evaluated their utility -- palatal structure, pelvic musculature, deep plantar tendons, intestinal convolutions, carotid arteries and the condition of the fifth secondary.  These characters vary among groups and the degree of congruence with one another and with other evidence of relationships is not consistent.  However, Gadow's careful and laborious comparisons were far more impressive than anything that had been done earlier and seemed to indicate that nothing better could be achieved.  In his 1893 treatise on the systematics of birds in Bronn's Klassen und Ordnungen des Thier-Reichs, Gadow reviewed many previous classifications, discussed each group of birds and presented his classification (pp. 299-303) (see Sibley and Ahlquist 1990: 222-224).

The sequence of the major groups in Gadow's classification begins with the ratites followed by the waterbirds.  The association of all waterbirds reflected the assumption that life began in the ocean and, therefore, waterbirds must be among the oldest or most "primitive" groups -- so the loons, grebes, penguins and other aquatic groups were placed near the beginning of the series in most classifications.
 In 1930 Alexander Wetmore adopted Gadow's classification with minor modifications, such as some ordinal names.  Wetmore took a few items from other authors, but Gadow was the main source.  Wetmore's version of Gadow's classification became the standard because there was no evidence that it was possible to do better with the methods available.

In 1931, the Wetmore classification was adopted by James L. Peters for his Checklist of Birds of the World.  This became a 15 volume series published between 1931 and 1962.
Ernst Mayr edited the later volumes after Peters' death in 1952.  Mayr abandoned the Wetmore classification for the oscine passerines and used a system known as the "Basel sequence" because it was developed by a committee at the 11th International Ornithological Congress in Basel, Switzerland in 1954 (see Sibley and Ahlquist 1990, pp. 632-634 for details).
These are not the only classifications of birds that have been proposed during the past century, but they illustrate the methods and problems associated with the process.
A classification must contain more information than a list of names and the most reasonable basis is a reflection of the phylogeny.