|Home | About | Journals | Submit | Contact Us | Français|
The standard nomenclature that has been used for many telencephalic and related brainstem structures in birds is based on flawed assumptions of homology to mammals. In particular, the outdated terminology implies that most of the avian telencephalon is a hypertrophied basal ganglia, when it is now clear that most of the avian telencephalon is neurochemically, hodologically, and functionally comparable to the mammalian neocortex, claustrum, and pallial amygdala (all of which derive from the pallial sector of the developing telencephalon). Recognizing that this promotes misunderstanding of the functional organization of avian brains and their evolutionary relationship to mammalian brains, avian brain specialists began discussions to rectify this problem, culminating in the Avian Brain Nomenclature Forum held at Duke University in July 2002, which approved a new terminology for avian telencephalon and some allied brainstem cell groups. Details of this new terminology are presented here, as is a rationale for each name change and evidence for any homologies implied by the new names.
Revisions for the brainstem focused on vocal control, catecholaminergic, cholinergic, and basal ganglia-related nuclei. For example, the Forum recognized that the hypoglossal nucleus had been incorrectly identified as the nucleus intermedius in the Karten and Hodos (1967) pigeon brain atlas, and what was identified as the hypoglossal nucleus in that atlas should instead be called the supraspinal nucleus. The locus ceruleus of this and other avian atlases was noted to consist of a caudal noradrenergic part homologous to the mammalian locus coeruleus and a rostral region corresponding to the mammalian A8 dopaminergic cell group. The midbrain dopaminergic cell group in birds known as the nucleus tegmenti pedunculopontinus pars compacta was recognized as homologous to the mammalian substantia nigra pars compacta and was renamed accordingly; a group of γ-aminobutyric acid (GABA)ergic neurons at the lateral edge of this region was identified as homologous to the mammalian substantia nigra pars reticulata and was also renamed accordingly. A field of cholinergic neurons in the rostral avian hindbrain was named the nucleus pedunculopontinus tegmenti, whereas the anterior nucleus of the ansa lenticularis in the avian diencephalon was renamed the subthalamic nucleus, both for their evident mammalian homologues.
For the basal (i.e., subpallial) telencephalon, the actual parts of the basal ganglia were given names reflecting their now evident homologues. For example, the lobus parolfactorius and paleostriatum augmentatum were acknowledged to make up the dorsal subdivision of the striatal part of the basal ganglia and were renamed as the medial and lateral striatum. The paleostriatum primitivum was recognized as homologous to the mammalian globus pallidus and renamed as such. Additionally, the rostroventral part of what was called the lobus parolfactorius was acknowledged as comparable to the mammalian nucleus accumbens, which, together with the olfactory tubercle, was noted to be part of the ventral striatum in birds. A ventral pallidum, a basal cholinergic cell group, and medial and lateral bed nuclei of the stria terminalis were also recognized.
The dorsal (i.e., pallial) telencephalic regions that had been erroneously named to reflect presumed homology to striatal parts of mammalian basal ganglia were renamed as part of the pallium, using prefixes that retain most established abbreviations, to maintain continuity with the outdated nomenclature. We concluded, however, that one-to-one (i.e., discrete) homologies with mammals are still uncertain for most of the telencephalic pallium in birds and thus the new pallial terminology is largely devoid of assumptions of one-to-one homologies with mammals. The sectors of the hyperstriatum composing the Wulst (i.e., the hyperstriatum accessorium intermedium, and dorsale), the hyperstriatum ventrale, the neostriatum, and the archistriatum have been renamed (respectively) the hyperpallium (hypertrophied pallium), the mesopallium (middle pallium), the nidopallium (nest pallium), and the arcopallium (arched pallium). The posterior part of the archistriatum has been renamed the posterior pallial amygdala, the nucleus taeniae recognized as part of the avian amygdala, and a region inferior to the posterior paleostriatum primitivum included as a subpallial part of the avian amygdala. The names of some of the laminae and fiber tracts were also changed to reflect current understanding of the location of pallial and subpallial sectors of the avian telencephalon. Notably, the lamina medularis dorsalis has been renamed the pallial-subpallial lamina. We urge all to use this new terminology, because we believe it will promote better communication among neuroscientists.
The various structures of the human brain were typically named for their shape, appearance, or position, and then these names were applied to the presumed homologous structures in nonhuman mammals, even if the descriptive implications were inappropriate for the given nonhuman species. For example, in rodents, the substantia nigra is not black and the dorsal lateral geniculate nucleus is not bent like a flexed knee, yet these terms have nonetheless been applied in rodents. The presumption has been that a communication benefit derives from calling homologous structures by the same name and that this benefit outweighs any descriptive or etymological incongruities. A similar benefit in parsimony and communication ensues if homologous structures are given the same names across classes. The issue of brain homology across classes was addressed, beginning more than 100 years ago, by Ludwig Edinger and his students in Germany, as part of an effort to understand vertebrate brain evolution (Edinger et al., 1903; Edinger, 1908). They, and later others such as J.B. Johnston, G.C. Huber, E.C. Crosby, C.U. Ariëns-Kappers, and C.J. Herrick, developed a theory of vertebrate brain evolution based on studies of cellstained and fiber-stained sections of fish, amphibian, reptilian, avian, and mammalian brains. This theory postulated that the brain had evolved by a stepwise addition of parts (reminiscent of Buffon’s Scala Naturae view of animal intelligence) and a gradual transfer of function from mesencephalic to ventral telencephalic, and ultimately to dorsal telencephalic centers (a process called encephalization; Buffon, 1749; Ariëns-Kappers, 1922, 1928; Johnston, 1923; Ariëns-Kappers et al., 1936; Lovejoy, 1936; Herrick, 1948, 1956).
Of present relevance is their interpretation of telencephalic evolution and the impact of this view on the terminology that came to be widely employed for the telencephalon of birds (Fig. 1A,B). In this view, the telencephalon of stem reptiles (now called stem amniotes) was thought to have consisted of a meager cortical mantle (the pallium) and a disproportionately large basal ganglia. In the evolution of stem amniotes into mammals, the meager cortical mantle was considered to have become enormously hypertrophied and elaborated into the neocortex. In contrast, in the lineage leading to birds, the basal ganglia was presumed to have grown in size and complexity, because a laminated structure resembling mammalian neocortex was not evident (Fig. 2A,C). The divergence between birds and mammals in the part of the telencephalon assumed to have elaborated was taken to explain the presumed differences in behavior between birds and mammals. Mammals were considered to possess a flexible, learned behavioral repertoire, and it was widely presumed that a neocortex was uniquely necessary for such abilities, whereas the complex behaviors of birds were commonly thought to be entirely instinctive, because it was thought that the basal ganglia controlled instinctive behavior (Elliot-Smith, 1901; Edinger, 1908; Herrick, 1956).
The specific view of telencephalic evolution that thus came to be widespread during the first 60 years of the 20th century was that both birds and mammals shared several basal ganglia structures, namely, an older structure inherited from fish called the paleostriatum primitivum (now called the globus pallidus in mammals) and a newer basal ganglia structure evolved in amphibians, but expanded in reptiles and more so in birds, called the neostriatum (considered equivalent to most of the caudate and putamen of mammals by Ariëns-Kappers et al. , but now used to refer to the entire caudatoputamen; Fig. 1A,B). Reptiles were thought to have also elaborated the two parts of the fish paleostriatum, the primitivum and the augmentatum (the latter considered by Ariëns-Kappers et al. to include primitive parts of the mammalian caudate and putamen) into two clearly distinct regions and to have passed this trait on to birds (Fig. 1A,B). Finally, the neostriatum in birds was thought to have given rise to a novel overlying structure called the hyperstriatum (Fig. 1A,B). Birds and mammals were also thought to share a caudobasal subcortical structure termed the archistriatum (Ariëns-Kappers, 1922, 1928), in mammals now called the amygdala. Although some investigators such as Kuhlenbeck (1938), Rose (1914), and Källén (1953) dissented from these views, the accretionary theory of vertebrate brain evolution, as espoused in major books by Ariëns-Kappers et al. (1936) and Herrick (1948, 1956), became the prevailing view and led to the predominant use of the terms neostriatum, archistriatum, and hyperstriatum to refer to the major telencephalic sectors above the paleostriatum in birds, and to the term neocortex for the major pallial sector of the telencephalon in mammals.
In 1967, Karten and Hodos published their stereotaxic atlas of the pigeon brain, which provided the first comprehensive effort to identify and name all parts of the brain in birds. In their Introduction, they describe the process by which they chose the terminology they used. For the hindbrain and midbrain, they adapted the terminology for the mammalian brainstem developed by Olszewski and coworkers (Meessen and Olzsewski, 1949; Olszewski and Baxter, 1954). For the pretectum and thalamus, they relied on the avian terminology of Kuhlenbeck (1937, 1939), Huber (1929), and Craigie (1931). For the telencephalon, they expressed concern that use of terms such as archistriatum, neostriatum, and hyperstriatum, as per Ariës-Kappers et al. (1936), could promote the notion that the telencephalon in birds is composed almost entirely of a highly developed basal ganglia, which they regarded as unlikely, based on developmental studies by Källén (1953) and Haefelfinger (1957). Although they noted that other telencephalic terminologies had been suggested for birds, such as the descriptive terminology of Kuhlenbeck (1938), or the letter-based terminology of Rose (1914), Karten and Hodos ultimately decided to employ the evolution-based nomenclature of Ariës-Kappers et al. (1936), which was a revision of Edinger’s terminology (Edinger et al., 1903; Edinger, 1908), for the major subdivisions of the telencephalon in birds, because this terminology was already entrenched.
Subsequent atlases for other avian species (Stokes et al., 1974; Youngren and Philips, 1983; Kuenzel and Masson, 1988; Matochik et al., 1991) largely used the same terminology as Karten and Hodos, leading to standardization of avian brain nomenclature. Although much of the terminology used in Karten and Hodos (1967) has stood the test of time, many interpretations of telencephalic homology implied by the terminology of Ariëns-Kappers et al. (1936) that Karten and Hodos (1967) had already suspected to be flawed have since been overwhelmingly shown to be erroneous (Figs. 1C, 2A–F). In particular, the notion that most of the telencephalon in birds is a hypertrophied basal ganglia has been disproved, because it is now clear that most of the telencephalon in birds is neurochemically, hodologically, and functionally comparable to the mammalian neocortex, claustrum, and pallial amygdala (all of which derive from the pallial sector of the developing telencephalon; Karten, 1969; Reiner et al., 1998a; Puelles et al., 2000). Additionally, the mammalian homologues of some brainstem cell groups connected with the telencephalon, which were not known at the time the Karten and Hodos atlas was completed, have been demonstrated. As deeper insight has been gained into the evolution, development, and function of avian and mammalian brains, it has become clear that the flawed homologies implied by the Ariëns-Kappers et al. (1936) terminology for the telencephalon of birds, as well as some now evident errors in brainstem terminology, make findings on avian brains confusing or inaccessible to researchers working on other species and perpetuate an outdated view of avian brain evolution, as well as vertebrate brain evolution in general.
This issue came to be of growing concern to avian neurobiologists, in part because of increased research focusing on avian brains. Formal efforts to revise avian brain nomenclature were begun 6 years ago by a small group of avian brain specialists. To develop widely acceptable new terms, this group sought to involve a greater number of researchers than had participated in two previous attempts to standardize avian neuroanatomical terms (Baumel, 1979, 1993). Accordingly, the group eventually grew to an international collection of multidisciplinary neuroscientists, and 2 years ago the group decided to hold an open Brain Nomenclature Forum, at which a new terminology would be adopted. This Forum was held July 18–20, 2002 at Duke University in Durham, North Carolina, and was attended by the authors of this paper. The Forum was preceded by in-depth discussions by E-mail and telephone of specific recommendations as to possible new terms. This manuscript presents the new terminology approved by the Forum, as well as the major evidence for the new terminology. A companion manuscript available on-line describes in more detail the pre-Forum preparatory period and the Forum logistics (http://www.interscience.wiley.com/jpages/0021-9967/suppmat/index.html). Details of the discussion and decisionmaking process will be published in a special issue of Brain, Behavior and Evolution devoted to the nomenclature change.
For each structure listed in the following sections, the old name is given first, followed by an arrow pointing to the new name. The old and new names for each structure are also shown in Tables 1-7, together with Latin and English versions of the new name and the abbreviations for each. The new names for several major telencephalic sectors are also schematized in Figure 1D. The old names and abbreviations used are those that are most commonly used in the published literature for each given structure. In instances in which the new avian name indicates homology to a given mammalian structure, the new names and abbreviations recommended by the Forum are those that are most commonly used in the mammalian literature. In the case of structures for which an entirely new name has been adopted by the Forum, terms close in form to the outdated names were typically selected (e.g., arcopallium as the new term to replace archistriatum). Abbreviations follow the rule that the first letters of the words represented in the abbreviation be capitalized, and only subordinate letters (e.g., second letter of a word) or letters representing subordinate words (such as the ventralis of pars ventralis) be in lower case.
The Forum focused on several brainstem cell groups connected with the basal ganglia or the song system (Table 1), for which the homology implied by the name was clearly incorrect, or at best obscure, and the true homologue is now amply demonstrated. These structures, new terms, and evidence for homology accepted by the Forum are detailed below on a structure-by-structure basis. The old and new names for these brainstem cell groups are also shown in Table 1.
In the Karten and Hodos pigeon brain atlas, a population of motoneurons located ventral to the dorsal motor nucleus of the vagus nerve and the nucleus intercalatus at levels straddling the obex was named the nucleus intermedius, following the practice of earlier workers (Ariëns-Kappers et al., 1936). Because this nucleus is now known to innervate tongue, tracheal, and syringeal muscles via the lingual and tracheosyringeal branches of the 12th nerve (Hillebrand, 1971; Nottebohm et al., 1976; Wild and Zeigler, 1980; Wild, 1981; Eden and Correia, 1982; Youngren and Phillips, 1983; Vicario and Nottebohm, 1988; Dubbeldam and Bout, 1990; Wild, 1990), the Nomenclature Forum decided to formally endorse the renaming of the nucleus intermedius of Karten and Hodos as the hypoglossal nucleus, or nXII (Table 1, Fig. 3A). The correct name for this cell group had already been routinely used for many years in various publications (e.g., Nottebohm et al., 1976; Medina and Reiner, 1994) but had not yet been formally adopted.
A population of motoneurons ventral to the nucleus intermedius of Karten and Hodos (1967) and abutting the lateral edge of the medial longitudinal fasciculus was identified in their atlas as the hypoglossal nucleus. This group of motoneurons is made up of dorsomedial and ventrolateral clusters, both of which are now known to innervate upper neck muscles, but not the tongue (Wild, 1981; Eden and Correia, 1982; Watanabe and Ohmori, 1988; Horster et al., 1990; Zijlstra and Dubbeldam, 1994). This cell group is separate from an upper cervical spinal cord nucleus from which the accessory nerve arises (Beccari, 1943; Hillebrand, 1975; Wild, 1981; Zijlstra and Dubbeldam, 1994). The Nomenclature Forum therefore decided to rename this cell group formally as the nucleus supraspinalis, the name that had been employed in several studies subsequent to the publication of the Karten and Hodos atlas (Hillebrand, 1975; Wild and Zeigler, 1980; Wild, 1981) (Table 1, Fig. 3A).
In the Karten and Hodos (1967) atlas, the locus ceruleus (or coeruleus) was identified as a large neuronal field extending from rostral pontine to caudal midbrain tegmental levels. Subsequent atlases in other avian species have followed this convention (Stokes et al., 1974; Youngren and Phillips, 1983; Kuenzel and Masson, 1988; Matochik et al., 1991). The LoC in mammals is a cell-dense nucleus containing neurons rich in noradrenaline (Moore and Card, 1984; Smeets and Reiner, 1994) that is restricted to rostral pontine levels and does not extend into the midbrain (Puelles and Medina, 1994). Immunohistochemical studies of various avian species over the past 15 years have shown that only the caudal part of what is identified as the LoC (i.e., the hindbrain part) in the Karten and Hodos atlas contains noradrenergic neurons (von Bartheld and Bothwell, 1992; Bailhache and Balthazart, 1993; Reiner et al., 1994; Mello et al., 1998a) and is in the same topographic and segmental location as mammalian LoC (Puelles and Medina, 1994). A neurochemically and topologically identical cell group is present in the rostral hindbrain of reptiles (turtles, lizards, and snakes), as well as urodele and anuran amphibians (Gonzalez and Smeets, 1994; Medina et al., 1994; Smeets, 1994), and is considered homologous to the caudal part of the Karten and Hodos LoC in birds and the LoC of mammals (Smeets and Reiner, 1994). Thus, the Forum redefined the LoC in birds as the noradrenergic cell field at the lateral edge of the fourth ventricle at rostral pontine levels (Table 1, Figs. 3B, 4A,B), with coeruleus rather than ceruleus being the recommended spelling due to its more common usage. It should be noted that for both birds and mammals the A6 cell group does not correspond exclusively to the cytoarchitectonically defined locus coeruleus, because the noradrenergic neurons of A6 extend somewhat ventral and rostral to the LoC (Dahlström and Fuxe, 1964; Hokfelt et al., 1984; von Bartheld and Bothwell, 1992; Reinger et al., 1994), and nearby fields of cholinergic and serotonergic neurons overlap somewhat with the LoC (Steinbusch, 1981; Mesulam et al., 1984; Steinbusch, 1984; Yamada et al., 1984; Vincent and Reiner, 1987; Coozi et al., 1991; Woolf, 1991; von Bartheld and Bothwell, 1992; Medina and Reiner, 1994). Nonetheless, the A6 largely coincides with the LoC, and noradrenergic neurons represent the predominant population in the LoC.
The cell group identified in Karten and Hodos (1967) as the rostral part of the LoC (i.e., the LoC at levels A2.75–A2.25 of the pigeon brain atlas) actually corresponds to the midbrain A8 dopaminergic cell group. Studies of various avian species over the past 15 years have established that neurons in the rostral (i.e., midbrain) part of what was termed the LoC in the Karten and Hodos atlas are dopaminergic rather than noradrenergic (Smeets, 1991; Bailhache and Balthazart, 1993; Reiner et al., 1994). They coincide topographically with the A8 cell group of mammals and show the same continuity with the SNc/A9 group as does the A8 of mammals. Developmental studies further reinforce the conclusion that the rostral part of what was called the LoC in the Karten and Hodos atlas is the A8 (Puelles and Medina, 1994). A distinct A8 with similar neurochemical and topographic traits as those of birds and mammals also is characteristic of reptiles (turtles, lizards, and snakes) and urodele and anuran amphibians (Gonzalez and Smeets, 1994; Smeets, 1994; Smeets and Reiner, 1994). Thus, the Forum renamed the rostral part of what had been called the LoC in birds as the A8 (Table 1, Figs. 3C, 4C,D). One further issue of concern was whether the avian A8 cell group should also be given the same common name as the A8 cell group in mammals, namely, the retrorubral nucleus. We decided that although this might be useful for further emphasizing the homology to mammals, the A8 in birds is not truly retrorubral in position, and the name was thus potentially misleading. Additionally, the A8 is but one of the cellular occupants of the retrorubral region in mammals (Hokfelt et al., 1984; Kitahama et al., 1994; Paxinos and Watson, 1998), and use of this name for birds would imply homologies beyond that to the A8 population, homologies that at this time have not been investigated.
Some areas of the brainstem in birds have names that produce confusion because of implied similarity to nonhomologous areas of the mammalian brainstem. This is the case for the nucleus tegmenti pedunculopontinus pars compacta, located in the diencephalic-mesencephalic tegmentum (Puelles and Medina, 1994). This cell field contains a large population of dopaminergic neurons that send a massive dopaminergic innervation to the striatal part of the basal ganglia in birds (i.e., the regions that have been called lobus parolfactorius and paleostriatum augmentatum; Kitt and Brauth, 1986a,b; Bailhache and Balthazart, 1993; Reiner et al., 1994; Szekely et al., 1994; Karle et al., 1996; Metzger et al., 1996; Durstewitz et al., 1999) and therefore is accepted as homologous to the substantia nigra pars compacta of other vertebrates (Puelles and Medina, 1994; Smeets and Reiner, 1994; Medina and Reiner, 1995; Reiner et al., 1998a). The name applied to this region in Karten and Hodos (1967) and in many subsequent atlases and published papers, however, incorrectly suggests a homology with the pedunculopontine tegmental nucleus of mammals, located in rhombomere 1. This rhombomere gives rise to the rostral hindbrain region that includes the LoC, and it is characterized by the presence of numerous cholinergic neurons but no dopaminergic neurons (Medina and Reiner, 1994; Puelles and Medina, 1994). Moreover, the actual avian homologue of the mammalian pedunculopontine tegmental nucleus, which contains cholinergic neurons, has been identified in rhombomere 1 (Figs. 3F, ,4E;4E; Medina and Reiner, 1994), as detailed below in the section on this nucleus. To rectify these misnomers and avoid confusion, the Forum renamed what had been called the nucleus tegmenti pedunculopontinus pars compacta (TPc) in birds as the substantia nigra pars compacta (SNc), or the A9 dopaminergic cell group (Table 1, Figs. 3D,F, 4F). Although the A9 dopaminergic neuron field in birds is not as compact as in rodents, or as pigmented as in humans, A9 varies in its compactness and blackness (i.e., pigmentedness) even among mammals. For this reason, and because of the gain in using a homology-based term for avian A9, the Forum decided that the descriptive inaccuracies of the terms “compacta” and “nigra” were far outweighed by the benefits obtained in adopting SNc as its common name.
In mammals, the substantia nigra pars compacta lies dorsal to a dense field of fibers and terminals that arise from the striatal part of the basal ganglia, in which are embedded scattered GABAergic neurons (Reiner et al., 1998a). Because of the reticulated appearance imparted to this region by its fiber richness, it is called the substantia nigra pars reticulata (SNr). The striatal terminals in mammalian SNr also are GABAergic, predominantly contain the neuropeptides substance P (SP) and dynorphin (DYN), and end on both the SNr GABAergic neurons and the dendrites of SNc dopaminergic neurons that extend into the SNr (Reiner et al., 1998a). The GABAergic neurons of the mammalian SNr project to the intermediate layers of the superior colliculus (which project to hindbrain premotor areas) and to the ventral anterior and ventrolateral thalamic nuclei (which project to motor cortices). By means of these projections, SNr neurons are involved in the basal ganglia-mediated control of movement (Reiner et al., 1998a). A similar GABAergic cell population has been identified in birds, lateral to and overlapping the SNc/A9 (Veenman and Reiner, 1994). This cell group was initially called the substantia nigra pars lateralis (Veenman and Reiner, 1994; Medina and Reiner, 1997), and it was shown to receive striatal input from SP/DYN-containing terminals of striatal origin and to project to the tectum and thalamus (Reiner et al., 1983; Anderson et al., 1991; Medina and Reiner, 1997; Reiner et al., 1998a). A similar cell group is also present in reptiles (turtles and crocodilians), supporting the homology of this cell group among amniotes (Brauth and Kitt, 1980; Reiner et al., 1980, 1998a; Brauth et al., 1983; Reiner and Carraway, 1987). The Forum thus concluded that the SNL of birds is homologous to the SNr in mammals and decided to rename avian SNL to SNr (Table 1, Figs. 3D,F, ,4F).4F). As with SNc, application of the term SNr to birds yields a slight descriptive misimpression, because this structure in birds is not as reticulated in appearance as in mammals and is not part of a field that contains black (i.e., pigmented) cells. The Forum decided that such inaccurate implications were outweighed by the benefits of adopting a homology-based term such as SNr
The area ventralis of Tsai of Karten and Hodos (1967) is considered homologous to the mammalian ventral tegmental area (Bailhache and Balthazart, 1993; Puelles and Medina, 1994, Reiner et al., 1994), which is also known as the A10 dopaminergic cell group and had once commonly been called the ventral tegmental area of Tsai (Crosby et al., 1962). As in mammals, the cell group in birds is dopaminergic, is situated near the oculomotor nerve as it exits the midbrain, is rostromedial to and continuous with the SNc/A9, and sends a massive dopaminergic projection to the basal ganglia, mainly to medial and ventral parts of the region that has been called the lobus parolfactorius (LPO; Lewis et al., 1981; Kitt and Brauth, 1986b, Bottjer, 1993; Reiner et al., 1994; Szekely et al., 1994). As for A9 and A8, a comparable cell group is present in all orders of reptiles (Smeets, 1994; Smeets and Reiner, 1994). For the purpose of abandoning the eponym Tsai (because eponyms are no longer employed according to international rules of anatomical nomenclature; Baumel, 1979, 1993) and formally adopting the term already employed for mammals and in many recent studies in birds, the Forum approved the renaming of the avian area ventralis of Tsai as the ventral tegmental area, with the acceptable alternative name of A10. (Table 1, Figs. 3D,F, ,4F4F).
In mammals, a cholinergic cell group is located in the tegmentum of rhombomere 1, caudal to A9 and A8, as is easily seen in sagittal sections (Mesulam et al., 1984; Woolf, 1991). This cell group has been termed the pedunculopontine tegmental nucleus (also sometimes abbreviated as PPN). A similarly situated cholinergic cell group has been identified in chicken and pigeon (von Bartheld and Bothwell, 1992; Medina and Reiner, 1994), a rostroventral part of which overlaps what was called the nucleus profundus mesencephali pars ventralis in Karten and Hodos (1967). This cholinergic cell group closely resembles the mammalian pedunculopontine tegmental nucleus in its projections to the midbrain tectum, retinorecipient thalamic and pretectal nuclei, and telencephalon (Reiner et al., 1982; Medina and Reiner, 1994). A comparably situated cholinergic cell group has been recognized in turtles and lizards (Medina et al., 1993; Powers and Reiner, 1993), supporting the homology of this cell group among amniotes. The Forum thus recognized the existence of this cell group in birds and recommended that the same name be employed as for its mammalian homologue (Table 1, Figs. 3F, ,4E4E).
The avian anterior nucleus of the ansa lenticularis is an inconspicuous cell group of the rostral diencephalon located in and along the medial edge of the ansa lenticularis (a fiber bundle interconnecting the basal ganglia with various brainstem cell groups). The ALa was not identified as such in Karten and Hodos (1967) and was first recognized as a projection target of the pallidal part of the basal ganglia by Karten and Dubbeldam (1973). Subsequent studies (Brauth et al., 1978; Medina and Reiner, 1995, 1997; Reiner et al., 1998a), culminating with Jiao et al. (2000), revealed the avian ALa to be a glutamatergic cell group that develops within the same prosencephalic region, is located in the same final adult position as the mammalian subthalamic nucleus, and is reciprocally connected with the pallidal part of the basal ganglia. In terms of its function, the neurotransmitters utilized by its inputs and outputs, its developmental profile, its position in the diencephalon, and its likely presence in reptiles, the ALa clearly appears homologous to the subthalamic nucleus of mammals (Reiner et al., 1998a; Jiao et al., 2000) . For this reason, the Forum concluded that the avian ALa should be renamed the subthalamic nucleus (STN; Table 1, Fig. 3E).
The basal ganglia in mammals forms within a ventral part of the developing telencephalon called the subpallium. The subpallium, which contains the septal nuclei and several other nuclei in addition to those of the basal ganglia, is notably distinct from the overlying telencephalic region called the pallium in its connectivity, neurochemistry, and the genes that regulate its development (Fig. 2B,D–F; Rubenstein et al., 1994; Reiner et al., 1998a; Swanson and Petrovich, 1998). Developmental, topological, neurochemical, cellular, connectional, and functional data all strongly support the conclusion that the subpallial region lateral to the telencephalic ventricle in birds and reptiles contains homologues of mammalian basal ganglia, whereas the subpallial region medial to the lateral ventricle in birds and reptiles contains the homologues of the mammalian septal nuclei (Karten and Dubbeldam, 1973; Brauth and Kitt, 1980; Brauth et al., 1983; Reiner et al., 1984a, 1998a; Smeets, 1994, Medina and Reiner, 1995, 1997; Medina et al., 1997; Smith-Fernandez et al., 1998; Puelles et al., 2000).
Embryological and developmental molecular studies in birds and mammals further indicate that the developing subpallium consists of two radially oriented histogenetic zones, a dorsal one that corresponds to the lateral ganglionic eminence of mammals and a more ventral one that corresponds to the medial ganglionic eminence of mammals (Puelles et al., 2000; Cobos et al., 2001a; Marin and Rubenstein, 2001; Redies et al., 2001). Among the derivatives of the dorsal zone (which expresses Dlx1/2 but not Nkx2.1) are the various striatal cell groups, which in mammals make up the dorsal striatum (i.e., the caudate and putamen) and the so-called ventral striatum (the nucleus accumbens and olfactory tubercle). The derivatives of the ventral zone (which expresses Dlx1/2 and Nkx2.1) in mammals include the dorsal pallidum (or globus pallidus) and the ventral pallidum. The dorsal striatal and pallidal cell groups are commonly said to make up the dorsal basal ganglia, or more simply the basal ganglia, whereas the ventral striatal and pallidal cell groups are said to make up the ventral (or limbic) basal ganglia (Heimer et al., 1985, 1997). These ventral regions differ from the basal ganglia proper in neuropeptide localization, connections, and function.
As with the brainstem, the homology implied by the existing name for a number of cell groups within the subpallium of birds was incorrect or at best obscure, and more recent data had established a consensus as to the correct homologue. The Forum thus sought to rename the various parts of the subpallium in birds to reflect more accurately their homologues in mammals. In particular, we sought to identify and rename the striatal and pallidal subdivisions of the avian dorsal basal ganglia, which as a whole is largely involved in somatomotor functions (Reiner et al., 1994; Doupe and Kuhl, 1999; Jiao et al., 2000), as well as those of the ventral basal ganglia. As in mammals, the striatum and pallidum can be distinguished because their neuron types differ in morphology, neuropeptide content, connections, and electrophysiology (Reiner et al., 1984a, 1998a; Reiner and Anderson, 1990; Medina and Reiner, 1995; Farries and Perkel, 1000). For example, the striatum is rich in dopaminergic fibers and acetylcholinesterase activity, whereas the pallidum is poor in both (Fig. 2B,E). The various subpallial structures, the new terms, and the evidence for homology accepted by the Nomenclature Forum are detailed below on a structure-by-structure basis and summarized in Table 2. Note that the Forum concluded that sufficient evidence existed to recognize a ventrocaudal part of the subpallium as part of the amygdala, as addressed in the section on the archistriatum.
Diverse lines of evidence indicate that the avian LPO has traits characteristic of the dorsal striatum in mammals. These include a prominent dopaminergic input (Fig. 2B) from the substantia nigra pars compacta (SNc/A9) and ventral tegmental area, enrichment in dopamine receptors, a projection back to SNc/A9 and VTA/A10, an acetylcholine-rich (Figs. 2E, ,5D)5D) and cholinesterase-rich neuropil, and an enrichment in GABAergic neurons that either contain SP/DYN (Fig. 5A–C) or enkephalin (Lewis et al., 1981; Bottjer, 1993; Casto and Ball, 1994; Grisham and Arnold, 1994; Reiner et al., 1994; Medina and Reiner, 1995; Soha et al., 1996; Reiner et al., 1998a; Luo and Perkel, 1999b; Sun and Reiner, 2000). Additionally, most of the LPO develops from a Dlx1/2-rich and Nkx2.1-poor neuroepithelium (Fig. 2F), as does the mammalian dorsal striatum (Smith-Fernandez et al., 1998; Puelles et al., 2000). For these reasons, and additional ones summarized in Reiner et al. (1998a), the Forum replaced the arcane name lobus parolfactorius (meaning lobe next to the olfactory bulb) with the term medial striatum and recognized it as (together with the paleostriatum augmentatum) forming the avian dorsal striatum (Table 2, Figs. 5A–F, 6A–D).
This change is attended by several qualifications. First, the medial striatum of birds does not appear to be homologous in a one-to-one fashion (also referred to as discrete homology by Smith, 1967) with the medial part of mammalian dorsal striatum, namely, the caudate nucleus. Principal among the reasons against such a notion is the difference between the LPO and caudate in their projections. In both birds and mammals, the striatum as a whole projects to the substantia nigra and the pallidum, with the projections arising from different populations of mediumsized neurons with heavily spiny dendrites (Gerfen, 1988, Graybiel, 1990; Reiner and Anderson, 1990; Reiner et al., 1998a). In mammals, such striatonigral and striatopallidal neurons are intermingled throughout the caudate and putamen (Beckstead and Cruz, 1986; Reiner and Anderson, 1990; Selemon and Goldman-Rakic, 1990; Reiner et al., 1998a). By contrast, the LPO of birds contains spiny neurons projecting to the substantia nigra but few (if any) that target the pallidal part of the dorsal basal ganglia, whereas the lateral striatum (paleostriatum augmentatum) projects primarily to the dorsal pallidum (Karten and Dubbeldam, 1973; Brauth et al., 1978; Reiner et al., 1983; Bottjer et al., 1989). Thus, wheras the medial striatum in birds (i.e., the LPO) does contain spiny projection neurons, it does not contain all of the same striatal projection neuron types as does the caudate in mammals.
Second, the sparse presence of a pallidal-type neuron projecting to the thalamus may be a unique feature of the avian medial striatum. This cell type was first discovered within a specialized cell group of the medial striatum in male songbirds called Area X (Fig. 6D). Although Area X predominantly contains the cell types typical of the mammalian striatum by physiology, dendritic morphology, and neurotransmitter features (Grisham and Arnold, 1994; Luo and Perkel, 1999b; Farries and Perkel, 2002b), it also contains a cell type that resembles a typical pallidal neuron in its aspiny morphology, its probable input from spiny striatal neurons, its GABAergic inhibitory projection to the thalamus, its neurochemistry, and its electrophysiology (Bottjer et al., 1989; Luo and Perkel, 1999a,b; Farries and Perkel, 2002b; Reiner et al., 2004). Scattered neurons with pallidal-type electrophysiology, connectivity, neurochemistry, and/or morphology have also been observed in the medial striatum proper surrounding Area X (Iyengar et al., 1999; Farries and Perkel, 2000; Reiner et al., 2004) and in the medial striatum proper of non-vocal learning birds such as chickens (Farries and Perkel, 2002a).
Moreover, some developmental studies have suggested that ventrolateral parts of the chicken medial striatum abutting the pallidum may ontogenetically be a pallidal territory that is heavily invaded by striatal cells during development and thereby becomes predominantly striatal in its cell-type composition (Puelles et al., 2000; Cobos et al., 2001a). If striatal cell-pallidal cell intermingling in part of the medial striatum proves to be a general avian trait that is absent from mammals, it might at that time be advisable to recognize some unique striatopallidal subdivision within the medial striatum and attach to it a specific suitable name. The Forum concluded, however, that insufficient data were available on the location of this region of mixing, on the prevalence of striatal and pallidal cell mixing in medial striatum across avian species, and on its absence from mammals to do so. Additionally, it was clear that what has been called the LPO has predominantly striatal cellular traits (Reiner et al., 1994, 1998a), and so it is appropriate for now to term it simply the medial striatum.
Although Area X of songbirds resides within the avian medial striatum (Nottebohm et al., 1976, Bottjer et al., 1989), its name is unaffected by the change in the name of LPO to medial striatum. Thus, the Forum recommended that Area X retain its name. A change to nucleus X was proposed, to reflect the clear boundaries of this structure; after discussion, the Forum took no position on whether Area X should be called nucleus X.
The Forum recognized that the change in the name of the LPO to the medial striatum does affect the name for the part of the budgerigar medial striatum involved in vocal control (Ball, 1994; Striedter, 1994; Durand et al., 1997), because LPO was part of the name for this structure. The Forum recommended that this region henceforth be called the magnocellular nucleus of the medial striatum.
Diverse lines of evidence indicate that the PA has striatal traits and, together with what has been called the LPO, constitutes the striatal part of avian dorsal basal ganglia (Karten, 1969; Lewis et al., 1981; Bottjer, 1993; Casto and Ball, 1994; Reiner et al., 1994, 1998a; Medina and Reiner, 1995; Soha et al., 1996; Farries and Perkel, 2000; Puelles et al., 2000; Sun and Reiner, 2000). The striatal traits of the PA include a prominent dopaminergic input (Fig. 2B) from the SNc/A9, an enrichment in dopamine receptors, an acetylcholine-rich (Figs. 2E, ,5D)5D) and cholinesterase-rich neuropil, an enrichment in GABAergic neurons that contain either SP/DYN (Fig.5A–C) or enkephalin, projections to the paleostriatum primitivum (now to be called the globus pallidus), and a glutamate receptor pattern quite similar to that of the mammalian striatum and avian medial striatum (Wada et al., 2001). Additionally, the PA develops from a Dlx1/2-rich and Nkx2.1-poor neuroepithelium (Fig. 2F; Smith-Fernandez et al., 1998; Puelles et al., 2000).
For these reasons, as well as reasons summarized in Reiner et al. (1998a), the Forum concluded that the name paleostriatum augmentatum (with its inaccurate evolutionary and cellular implications of being a pallidal derivative) should be abandoned and replaced with the term lateral striatum (Table 2, Figs. 5A–F, 6A–C). As for the avian medial striatum compared with the mammalian caudate nucleus, there is no compelling evidence that the lateral striatum of birds is homologous in a one-to-one fashion with the lateral part of the mammalian striatum, namely, the putamen. For example, the LSt in birds contains spiny neurons that mainly project to the pallidal part of basal ganglia but not to the substantia nigra (Karten and Dubbeldam, 1973; Anderson and Reiner, 1991a; Veenman et al., 1995a; Medina et al., 1997; Mezey and Csillag, 2002), whereas striatonigral and striatopallidal neurons are intermingled throughout the putamen (Beckstead and Cruz, 1986; Reiner and Anderson, 1990; Selemon and Goldman-Rakic, 1990; Reiner et al., 1998a).
The Forum recommended that the paleostriatum primitivum henceforth be called the globus pallidus (Table 2, Figs. 5A–F, 6B,C). This term is appropriate for descriptive reasons (both avian GP and its mammalian counterpart are pale) and because of neurochemical, cellular, hodological, phylogenetic, and developmental evidence that the avian GP is homologous to the mammalian GP (Karten and Dubbeldam, 1973; Brauth et al., 1978; Kitt and Brauth, 1981; Reiner et al., 1984a, 1998a; Reiner and Carraway, 1987; Veenman and Reiner, 1994; Medina and Reiner, 1995, 1997; Medina et al., 1997; Marin et al., 1998; Puelles et al., 2000; Cobos et al., 2001a; Gonzalez et al., 2002; Brox et al., 2003). In both birds and mammals, the projection neurons of the globus pallidus possess large cell bodies and smooth dendrites, derive from an Nkx2.1+ neuroepithelium, and give rise to the motor output projections of the basal ganglia. Additionally, globus pallidus neurons in both birds and mammals are GABAergic, contain the neuropeptide LANT6 (Fig. 5E), and receive inputs with a woolly fiber morphology from either SP/DYN-containing (Fig. 5A–C) or ENK-containing striatal neurons, as well as a glutamatergic input from the subthalamic nucleus (Reiner et al., 1998a, 1999; Jiao et al., 2000). In addition, functional evidence points to similarities in electrophysiological properties (Farries and Perkel, 2000). Avian globus pallidus neurons, however, appear to migrate farther laterally than do mammalian pallidal neurons (Puelles et al., 2000; Cobos et al., 2001a). Although the avian globus pallidus as a field is not as globular as the mammalian GP, even different mammalian species show variation in the shape of the GP, and the advantages of using globus pallidus as the new name for the PP in birds outweighs any slightly misleading implication as to its shape.
This enigmatic cell group is located within the lateral forebrain bundle below the inferior margin of the avian globus pallidus and is named for its location within the peduncle (i.e., forebrain bundle). Although Karten and Dubbeldam (1973) thought that its position resembled that of the mammalian internal pallidal segment, subsequent immunolabeling studies showed that it lacked the pallidal-type neurons and the SP/DYN-containing striatal input characteristic of the internal pallidal segment (Reiner et al., 1983; Reiner and Carraway, 1987; Anderson and Reiner, 1990; Veenman and Reiner, 1994). More recent studies have shown that the INP contains spiny neurons that express DARPP32, a striatal characteristic (Schnabel et al., 1997; Reiner et al., 1998b), a very similar glutamate receptor profile to the striatum (Wada et al., 2001), and numerous cholinergic neurons (Medina and Reiner, 1994), which is characteristic of, but not unique to, the striatum. Moreover, the INP has recently been found to develop within the striatal sector of the subpallium, to show continuity with the medial striatum, and to contain neurons expressing cell-surface proteins characteristic of striatal neurons (Medina et al., 2002). Although the data are thus beginning to suggest a largely striatal nature for the INP, the Forum concluded that it was still premature to conclude unequivocally that the INP is a striatal territory. Thus, the Forum decided to leave the name for the INP unaltered (Table 2, Figs. 5D, ,6B6B).
In the Karten and Hodos atlas, a region at the ventral tip of the lateral ventricle extending several millimeters rostral from the level of the anterior commissure, noticeably associated with a small bulge into the lateral ventricle, is identified as the nucleus accumbens. Other brain atlases of birds have followed this practice (Stokes et al., 1974; Youngren and Phillips, 1983; Kuenzel and Masson, 1988). As discussed in detail below, this region in actuality appears to be homologous to the lateral part of the mammalian bed nucleus of the stria terminalis. The nucleus accumbens in mammals is a ventral striatal territory that surrounds the inferior margin of the lateral ventricle at levels well anterior to the anterior commissure and is cytoarchitectonically continuous with the overlying caudatoputamen (i.e., the dorsal striatum) and with the olfactory tubercle below it (Heimer et al., 1985, 1997).
Although the hodological and neurochemical traits of mammalian nucleus accumbens grade into those of the overlying caudatoputamen, the accumbens and olfactory tubercle do differ from the dorsal striatum (Heimer et al., 1985, 1997). In particular, the accumbens is distinguished by its preferential reciprocal connections with the ventral tegmental area and its afferent input from limbic pallial regions such as the hippocampal complex, amygdala, and cingulate cortex, as well as from the frontal neocortex (Heimer et al., 1985, 1997; Paxinos and Watson, 1998). By contrast, the caudatoputamen has preferential reciprocal connections with the substantia nigra and receives its major pallial inputs from the neocortex (Heimer et al., 1985, 1997). By its cellular neurochemistry, the ventral striatum (including nucleus accumbens) also differs from the caudatoputamen. For example, spiny projection neurons of the caudate and putamen contain either enkephalin or both SP and DYN, but rarely co-contain SP and enkephalin (Anderson and Reiner, 1990; Besson et al., 1990; Reiner and Anderson, 1990; Lee et al., 1997). By contrast, SP co-localization with enkephalin in spiny projection neurons is much more common in the ventral striatum, including the nucleus accumbens (Anderson and Reiner, 1990; Furuta et al., 2002).
The region surrounding the tip of the lateral ventricle, at the ventromedial margin of the rostral pole of what has been called the LPO in birds, possesses the same topographic, hodological, and neurochemical traits as the nucleus accumbens of mammals (Brauth et al., 1978; Anderson and Reiner, 1990, 1991a; Reiner and Anderson, 1990; Veenman et al., 1995a; Mezey and Csillag, 2002; Roberts et al., 2002). By contrast, much of the remainder of what has been called the LPO is reciprocally connected with the substantia nigra pars compacta, receives pallial input from somatosensory and somatomotor areas of the pallium, and shows little co-localization of SP and enkephalin in spiny striatal projection neurons (Brauth et al., 1978; Reiner et al., 1983, 1984a,b; Anderson and Reiner, 1990, 1991b; Reiner and Anderson, 1990; Veenman et al., 1995b; Mezey and Csillag, 2002; Roberts et al., 2002). Thus, the region at the rostral ventromedial tip of medial striatum in birds is comparable to the mammalian nucleus accumbens. Moreover, a topographically, hodologically, and neurochemically similar cell group has been identified as the nucleus accumbens in turtles, lizards, and snakes (Russchen et al., 1987; Russchen and Jonker, 1988; Anderson and Reiner, 1990; Reiner and Anderson, 1990; Smeets, 1994; Guirado et al., 1999; Smeets et al., 2001).
For these reasons, the Forum recognized and recommended that the rostral ventromedial part of the former LPO of birds be called the nucleus accumbens and that the term medial striatum only be used to refer to the remainder of the LPO (Table 2, Fig. 6A,D). As in mammals, however, a precise cytoarchitectonic border between the dorsal striatum and nucleus accumbens is not evident, and a neurochemical criterion by which to distinguish the two fields unambiguously has not been identified. The finding that the medialmost sector of the medial striatum is reciprocally connected with the VTA and receives limbic pallial input (Veenman et al., 1995a; Mezey and Csillag, 2002) raises the possibility that the avian accumbens might extend more dorsally than shown in our figures (Fig. 6A,D). Additionally, whereas the nucleus accumbens of mammals possesses core and shell subdivisions, comparable subdivisions of the nucleus accumbens in birds have not been conclusively identified (Heimer et al., 1997; Roberts et al., 2002).
The olfactory tubercle is a telencephalic territory at the lower edge of the subpallium that was recognized in pigeons by Karten and Hodos (1973) and by most subsequent published papers or atlases dealing with the telencephalon in birds (Stokes et al., 1974; Youngren and Phillips, 1983; Kuenzel and Masson, 1988; Matochik et al., 1991). It is now established that, as in mammals, this cell group is a ventral striatal region that receives olfactory bulb input and resembles the olfactory tubercle of mammals in its neurochemistry and connectivity (Heimer et al., 1985, Heimer et al., 1997; Reiner and Karten, 1985; Reiner et al., 1994, 1998a; Roberts et al., 2002). The Forum thus endorsed the previously recognized homology of this cell group to the similarly named mammalian cell group and recommended no name change (Table 2, Figs. 6A,D, ,8B).8B). The Forum recommends a slight modification of the abbreviation for the olfactory tubercle (i.e., TuO) so it is not in conflict with the common abbreviation for the optic tract (i.e., TO).
A group of GABAergic neurons within the medial forebrain bundle (also called the fasciculus prosencephali medialis [FPM]) has been demonstrated in birds (Veenman and Reiner, 1994). This cell group is pallidal in nature because its glutamate receptor expression profile is identical to that of the GP (Wada et al., 2001) and because its neurons arise from the same Nkx2.1-expressing histogenetic subpallial neuroepithelial domain as the GP (Puelles et al., 2000). Additionally, it has the cellular neurochemistry, receives the ventral striatal inputs, and has the outputs characteristic of the ventral pallidum of mammals (Kitt and Brauth, 1981; Reiner and Carraway, 1987; Reiner and Anderson, 1990; Veenman and Reiner, 1994; Medina and Reiner, 1997; Reiner et al., 1998a; Roberts et al., 2002). A comparable cell group is present in turtles, crocodilians, and lizards (Brauth and Kitt, 1980; Brauth, 1984; Reiner, 1987; Reiner and Carraway, 1987; Russchen et al., 1987; Russchen and Jonker, 1988). The Forum thus recognized this cell group, which has sometimes been called the ventral paleostriatum (Kuenzel and Masson, 1988), and recommended it be referred to as the ventral pallidum (Table 2, Figs. 5C,D, ,6B).6B). The word ventral is used because this provides the VP with a name that is positionally appropriate with respect to its more dorsal counterpart, the GP, which has also been termed the dorsal pallidum.
The region identified as the nucleus accumbens by Karten and Hodos (1967) does not correspond to the nucleus accumbens of mammals, as was noted above, because it lacks the neurochemistry and cell types characteristic of the mammalian nucleus accumbens (Reiner et al., 1983, 1984a,b; Berk, 1987; Aste et al., 1998a,b). Instead, the Forum concluded that the area identified as the nucleus accumbens in Karten and Hodos is homologous to the lateral part of the mammalian bed nucleus of the stria terminalis, based on neurochemical and hodological evidence summarized by Aste et al. (1998a,b). The BSTL in birds has a rostral extension that tapers as it reaches the frontal pole of the medial striatum. This extension occupies a small but conspicuous bulge into the lateral edge of the inferior aspect of the telencephalic ventricle, and it may be surrounded by the true nucleus accumbens at very rostral levels (Reiner et al., 1983, 1984b; Aste et al., 1998a,b). The Forum thus recommended that the term nucleus accumbens be discontinued as the name for the region identified as accumbens in Karten and Hodos (1967) and that the lateral part of the bed nucleus of the stria terminalis be employed instead (Table 2, Figs. 5C,D,F, 6B,C). Note that the name for the BSTL in mammals (as well as that for the medial part of the bed nucleus of the stria terminalis) derives from the fact that it is embedded in a part of the stria terminalis (the dorsal part) and is thus a major target of this fiber bundle (De Olmos and Ingram, 1972). The stria terminalis in mammals has separate components that arise from diverse amygdaloid structures (Krettek and Price, 1977; Swanson and Petrovich, 1998), and the main fiber bundle itself courses from the amygdala toward the lateral edge of the anterior commissure, before ramifying into its ascending, commissural, and descending components (De Olmos and Ingram, 1972). Adopting the term BSTL for birds implies that a stria terminalis exists in birds and has the BSTL as one of its targets. The stria terminalis in birds appears to be at least partly represented by the hypothalamic part of the occipitomesencephalic tract, which courses medially from the archistriatum and taenia toward the lateral edge of the anterior commissure, before dividing into components that distribute to the BSTL and components that descend to the hypothalamus (Zeier and Karten, 1971).
The cell group located around the anterior commissure at the subpallio-preoptic transition in birds corresponds to the medial part of the mammalian bed nucleus of the stria terminalis, based on numerous topographic, hodological, and neurochemical criteria (Berk, 1987; Aste et al., 1998a,b; Jurkevich et al., 1999). This region in birds lies within the path of the apparent avian counterpart of the stria terminalis (Zeier and Karten, 1971). The Forum thus recommended that the term medial part of the bed nucleus of the stria terminalis be employed for this region in birds (Table 2, Fig. 6C). By in situ hybridization histochemistry and immunocytochemistry for arginine vasotocin, a single BSTM nucleus has been identified in Japanese quail (Aste et al., 1998a,b), whereas two BSTM subnuclei have been identified in chickens (Jurkevich et al., 1999). In chicken, the two BSTM subnuclei can be given the abbreviations BSTM1 and BSTM2, with formal names that follow suit. The BSTM in quail corresponds to the BSTM1 in chickens. It is important to note that the medial BST nuclei have been shown to be sexually dimorphic in chicken and quail and thus may be so in other avian species as well.
A field of large cholinergic neurons invests the lateral and medial forebrain bundles, the globus pallidus and ventral pallidum in birds (Medina and Reiner, 1994). This field is comparable to the field of large cholinergic neurons that spans the substantia innominata, ventral pallidum, nucleus basalis of Meynert, and globus pallidus in mammals (Medina and Reiner, 1994). In both birds and mammals, these cholinergic neurons innervate the pallium (Woolf, 1991; Medina et al., 1994), and a comparable cell group has been identified in lizards and turtles (Medina et al., 1993; Powers and Reiner, 1993). The Forum thus recommended that this field of cholinergic neurons in birds be named the nucleus basalis magnocellularis (Table 2, Figs. 5D, ,6B6B).
The Forum recognized, as proposed by Medina and Reiner (1994) based on their overview of data in diverse mammalian, avian, and reptilian species, that the large cholinergic neurons near the fasciculus diagonalis Brocae (in the septal area and medial basal forebrain), and surrounding the septomesencephalic tract (the latter renamed here as the septopallio-mesencephalic tract; see below), are homologous to the nucleus of the diagonal band of Broca in mammals and recommended that this name be used to refer to these neurons in birds (Table 2, Fig. 6B).
Available hodological data on input from the hippocampal complex and output to the hypothalamus (Krayniak and Siegel, 1978b; Atoji et al., 2002) suggest homology of the medial septal nucleus of birds with the nucleus of the same name in mammals. A comparable cell group is present in members of all reptilian orders (Brauth and Kitt, 1980; Brauth, 1984; Reiner and Carraway, 1987; Medina et al., 1993; Smeets, 1994). The Forum thus recommended that the name for the medial septal nucleus remain unchanged (Table 2, Fig. 6C).
Available hodological data on input from the hippocampal complex and output to the hypothalamus (Krayniak and Siegel, 1978b; Atoji et al., 2002), as well as neurochemical data (Reiner et al., 1984b, Reiner, 1994), suggest homology of the lateral septal nucleus of birds with the nucleus of the same name in mammals. A comparable cell group is present in all reptilian orders (Brauth, 1984; Reiner, 1987; Smeets, 1994). The Forum thus recommended that the name for the lateral septal nucleus remain unchanged (Table 2, Figs. 5C,D, 6B,C).
An overwhelming body of data now supports the conclusion that the dorsal three-fourths of the cerebrum in birds (including what has been termed the neostriatum, hyperstriatum, and archistriatum) is pallial in nature and therefore homologous as a field to the brain region of mammals that includes the neocortex, claustrum, and pallial amygdala (Karten, 1969, 1991; Güntürkün, 1991; Wild et al., 1993; Butler, 1994; Veenman et al., 1995b; Striedter, 1997; Smith-Fernandez et al., 1998; Medina and Reiner, 2000; Puelles et al., 2000). Our goal in revising the pallial terminology for birds was thus to replace names that possessed incorrect implications of homology to parts of the basal ganglia (e.g., the neostriatum, hyperstriatum, and archistriatum) with names that possessed correct implications of homology, or at the very least not incorrect ones. Although additional pallial regions such as the hippocampus and piriform cortex also received consideration, these did not require name change, and relatively little discussion was devoted to them. Discussion of the evidence for one-to-one homology between specific parts of the avian pallium bearing “-striatum” in their name and specific parts of the pallium in mammals (Karten and Shimizu, 1989; Bruce and Neary, 1995; Striedter, 1997; Smith-Fernandez et al., 1998; Puelles et al., 2000; Reiner, 2000; Butler and Molnar, 2002; Butler et al., 2002) led to the conclusion that one-to-one homologies were not established with certainty, and in some cases might not exist, due to divergent evolution between birds and mammals. This led then to the conclusion that the simplest course of action might be to replace “-striatum” with “-pallium” in the case of the above-noted structures and then devise some additional suitable prefixes or adjectives that distinguished the structures. Details of the possibilities considered will be presented in a special edition of Brain, Behavior and Evolution. In considering various choices, the Forum dealt with the new names for the neostriatum and hyperstriatum as a set, due to the relatedness of at least parts of these structures, whereas the archistriatal choices were considered separately. The results of the discussion follow.
In the existing terminology at the time of the Forum, the hyperstriatum ventrale (HV) and Wulst subregions (termed the hyperstriatum accessorium, hyperstriatum intercalatus superior, and hyperstriatum dorsale in the outdated nomenclature) were all named as parts of the hyperstriatum, but it has been clear for some time from developmental, hodological, neurochemical, and functional studies that they should not be linked in their names (Fig. 7A–D; Källén, 1953; Wächtler, 1985; Brauth et al., 1986; Wächtler and Ebinger, 1989; Csillag et al., 1993; Hodos, 1993; Shimizu et al., 1995; Husband and Shimizu, 1999; Denisenko-Nehrbass et al., 2000; Medina and Reiner, 2000; Sun and Reiner, 2000; Wada et al., 2001). The option approved by the Forum was to replace the term hyperstriatum in the various layers of Wulst with the term hyperpallium, to replace HV with mesopallium, and to replace neostriatum with nidopallium. These new terms are descriptive, because each prefix refers to an aspect of the relative degree of development or the location of the cell field to which it refers. For example, “hyper-” refers to an enlarged entity, which seems appropriate for the Wulst, because it is an enlarged (bulging) structure at the upper aspect of the pallium that represents a hypertrophied form of the dorsal cortex in reptiles (Medina and Reiner, 2000). A further advantage of the prefix “hyper-” is that it has already been used in the outdated name for the subdivisions of the Wulst. Thus, the term hyperpallium offers the benefit of easily linking the new term to the old, with abbreviations retained. The prefix “nido-”, which means nest, was considered apt for the neostriatum, because it is the pallial structure in which the overlying pallial structures are nested. Moreover, “nido-” allows abbreviation retention for the subregions within the nidopallium, and its similarity to the outdated prefix for this region (i.e., “neo-’) will facilitate learning the new term. Of course, the use of mesopallium (middle pallium) as a replacement for HV means that abbreviations need to change in this case. This was not seen as a serious disadvantage, however, because few subregions in mesopallium have been identified and named.
The named subdivisions of the Wulst have been referred to as pseudolayers (Medina and Reiner, 2000), because they seem to have some of the properties of cortical layers, except for pyramidal neurons with translaminar dendritic trees. Separate rationales for the renaming of the individual pseudolayers of the Wulst are discussed below. It needs to be noted that this name change does not affect the name of the individual functional areas (i.e., S1/M1 and V1) making up the Wulst, which each appear to span the depth of the various Wulst pseudolayers (Karten et al., 1973; Medina and Reiner, 2000; Wild and Williams, 2000; Table 3, Figs. 7, ,88).
The HA gives rise to the extratelencephalic, as well as some intratelencephalic, projections of the Wulst (Karten et al., 1973; Reiner and Karten, 1983; Wild, 1992; Shimizu et al., 1995; Veenman et al., 1995b; Kröer and Gürkün, 1999; Wild and Williams, 1999, 2000). The Forum decided to replace ”accessorium” with “apicale” because of the evolutionary or functional subordination to other parts of the hyperpallium implied by the term “acccessorium.” The term “apicale,” by contrast, more accurately describes the location of the HA (at the summit of the telencephalon) and allows abbreviation retention as well (Table 3, Figs. 6D, ,7A,7A, 8A,B). Along the edge of the HA flanking the region that has been called the hyperstriatum intercalatus superior is a band of small, densely packed neurons that Karten et al. (1973) named the intercalated nucleus of the HA (IHA). This region receives visual and somatosensory inputs from retinorecipient and dorsal column nuclear-recipient parts of the thalamus and relays to the HA (Karten et al., 1973; Miceli and Reperant, 1985; Wild, 1987, 1989, 1997; Funke, 1989a,b; Korzeniewska and Güntürkün, 1990; Güntürkün, 1991; Medina and Reiner, 2000). With the change of the name of the HA to the hyperpallium apicale, we recommend that the IHA be referred to as the interstitial nucleus of the HA, and retain the same abbreviation. The Forum recommended using interstitial rather than intercalated in the name for the IHA for two reasons. First, the word interstitial more accurately describes the relation of this cell field to the HA than does the word intercalated, which implies it is intercalated between the HA and another region. Second, to call it the intercalated nucleus promotes confusion with the second main pseudolayer of the Wulst, namely, the structure that has been called the hyperstriatum intercalatus superior.
This Wulst pseudolayer was identified in Karten and Hodos (1967) and subsequent atlases (Stokes et al., 1974; Youngren and Phillips, 1978; Kuenzel and Masson, 1988; Matochik et al., 1991). The name indicated that this region is intercalated between two others within what was called the hyperstriatum, namely. the HA and the underlying Wulst region previously called the hyperstriatum dorsale. The Forum recommends that the region formerly known as the hyperstriatum intercalatus superior now be termed the hyperpallium intercalatum (Table 3, Figs. 7A, 8A,B). The Forum decided to delete the “superior” from the name and the “S” from the abbreviation, because there is no hyperpallium intercalatus inferior. Note that switching interstitial for intercalatus in the name for the IHA eliminates any confusion between the IHA and HI (both had intercalatus in their prior names), and eliminating “superior” from the name for the HI avoids giving the impression that it should reside superior to either the IHA or HA.
In eliminating the term hyperstriatum ventrale, the Forum eliminated a valid basis for retaining “dorsale” in the name for the HD. To retain the abbreviation HD, the Forum approved the term hyperpallium densocellulare (Table 3, Figs 6D, ,7A,7A, 8A,B). The adjective “densocellular” has been used previously in the neuroanatomical literature (e.g., Jiminez-Castellanos and Graybiel, 1987; Popken et al., 2000; Fudge and Haber, 2001), and it appropriately describes the HD, because it stands out as cell dense in Nissl-stained material (Karten and Hodos, 1967).
The upper division of what was called the HV was referred to as its dorsoventral part (Karten and Hodos, 1967). This subdivision is distinct cytoarchitectonically (Rehkämper et al., 1984) and neurochemically from the lower part of the mesopallium (Fig. 7B,C; Wächtler, 1985; Faber et al., 1989; Wächtler and Ebinger, 1989; Csillag et al., 1993; Montagnese et al., 1993; Sun and Reiner, 2000). The term “dorsoventrale” was deemed needlessly complex, as already noted in Kuenzel and Masson (1988), and so the subordinate part of the name for this region has been simplified from that in Karten and Hodos (1967) to “dorsale.”
The lower division of the old HV was called its ventroventral part (Karten and Hodos, 1967). As noted, this subdivision is distinct cytoarchitectonically (Rehkämper et al., 1984) and neurochemically from the upper part of the mesopallium (Fig. 7B,C; Wächtler, 1985; Faber et al., 1989; Wächtler and Ebinger, 1989; Csillag et al., 1993; Montagnese et al., 1993; Sun and Reiner, 2000). The term “ventroventrale” was also deemed needlessly complex, as already noted in Kuenzel and Masson (1988), and so the subordinate name for this region has been simplified from that in Karten and Hodos (1967) to “ventrale”.
The oval nucleus of the HV is a distinct region recognized in parrots that is part of their telencephalic vocal control circuit (Striedter, 1994; Durand et al., 1997). A similar region has also been observed in songbirds and hummingbirds (Jarvis et al., 1998, 2000). With the renaming of the HV, this region becomes the oval nucleus of the mesopallium.
The caudal medial HV is a region recognized in guinea fowl, chicken, pigeons, songbirds, hummingbirds, and parrots as part of their telencephalic auditory circuit (Bonke et al., 1979a; Heil and Scheich, 1991a,b; Wild et al., 1993; Vates et al., 1996; Jarvis et al., 2000. With the renaming of HV, this region becomes the caudal medial mesopallium.
The region termed the IMHV has been recognized as a functionally and neurochemically definable cell field in studies of filial imprinting and passive avoidance learning in chicks (Horn, 1985; Rose and Csillag, 1985; Patel et al., 1988; Csillag et al., 1993). With the renaming of the HV, this region becomes the intermediate medial mesopallium.
The new term nidopallium maintains existing abbreviations, which is a considerable advantage given the large number of subregions within it (Rehkämper et al., 1985) that have names built upon the outgoing term neostriatum (Table 4, Figs. 7D, 8A–D). For example, the frontal, intermediate, and caudal neostriatum become the frontal, intermediate, and caudal nidopallium, and abbreviations remain unchanged. Because most papers and figures mainly use abbreviations to refer to specific structures, abbreviation retention for nidopallial structures will facilitate linking the old and new terms.
The term “neostriatum” should also be discarded in mammals, but for somewhat different reasons than in birds. In mammals, the term “neostriatum” is used by many investigators to refer to the caudatoputamen, but the term “striatum” is also used to refer to these very same structures (Carpenter, 1976). Thus the “neo-” in “neostriatum” is superfluous for mammals. In addition, the term “neostriatum” in mammals is only useful if there is some region that is old striatum (i.e., paleostriatum). Although the term “paleostriatum” was once used to refer to the globus pallidus in mammals (Carpenter, 1976), the term was abandoned long ago, and the assumption that the pallidum is evolutionarily older than the striatum has been disproved (Reiner and Carraway, 1985; Northcutt et al., 1988; Reiner et al., 1998a). Continued use of terms such as paleostriatum and neostriatum to refer to parts of the mammalian basal ganglia thus perpetuates discredited ideas about the evolution of the basal ganglia.
Three major nidopallial cell groups receive extratelencephalic sensory input in birds. These cell groups have been known as the ectostriatum (which receives visual input from nucleus rotundus of the thalamus; Karten and Revzin, 1966; Karten, 1969; Karten and Hodos, 1970), Field L (which receives auditory input from the nucleus ovoidalis of the thalamus; Karten, 1968, 1969; Wild et al., 1993), and the nucleus basalis (which receives trigeminal input directly from the principal sensory nucleus of the pons; Witkovsky et al., 1973). The ectostriatum and nucleus basalis but not Field L (Table 3, Figs. 7D, 8A–D) were renamed by the Forum for reasons discussed below. Each of these three regions consists of a core, which receives the majority of the ascending sensory input, and a surrounding shell, which mainly receives input from the core. The terminology that has been used for the shell region is problematic.
For example, Field L is often taken to mean the region in the caudomedial neostriatum (now the nidopallium) defined by Rose (1914) and shown by Karten (1969) in pigeons, and later in other avian species (Kelley and Nottebohm, 1979; Brauth et al., 1987; Vates et al., 1996), to receive a prominent input from nucleus ovoidalis. The work of Scheich and colleagues (e.g., Bonke et al., 1979a,b; Müller and Scheich, 1985) led to recognition that the auditory field in the caudomedial nidopallium was actually larger than the ovoidalis-recipient Field L alone. Thus, the ovoidalis- thalamorecipient zone was named L2 (Fig. 7D), and the regions immediately adjacent to L2, which receive L2 input as well as thalamic input from the periovoidalis region, were named L1 and L3. In guinea fowl, chicken, and pigeon, L1 is dorsomedial and L3 is ventrolateral to L2 (Bonke et al., 1979a; Heil and Scheich, 1985; Müller and Scheich, 1985; Wild et al., 1993). In songbirds, however, L1 is rostrodorsal, and L3 is caudoventral to L2 (Müller and Scheich, 1985; Fortune and Margoliash, 1992; Vates et al., 1996). As a consequence of the presence of subfields, the term Field L has come to have two different uses in the literature, one in which it refers to L2 alone and one in which it refers to L1–3.
Similar problems arise with respect to the ectostriatum, because in some studies the core and shell are together referred to as the ectostriatum, whereas in others only the thalamorecipient core is called the ectostriatum (Karten and Revzin, 1966; Karten and Hodos, 1970; Benowitz and Karten, 1976; Nixdorf and Bischof, 1982; Husband and Shimizu, 1999). Because the structure that has been known as the nucleus basalis also shows a core and shell organization, a similar problem exists for it (Fig. 7D; Wild and Zeigler, 1980; Wild et al., 1985; Veenman and Gottschaldt, 1986; Dubbeldam and Visser, 1987; Wild and Farabaugh, 1996). The Forum concluded that it would be desirable to develop a uniform and consistent terminology for these three sensory areas in the nidopallium and will make recommendations in a separate publication specifically devoted to this issue, in the special nomenclature issue of Brain, Behavior and Evolution.
The caudolateral neostriatum has been reported to possess neurochemical and functional similarities to the prefrontal cortex in mammals (Mogensen and Divac, 1982; Reiner, 1986). Although these similarities are thought to represent an instance of analogy due to parallel evolution, nonetheless, as a consequence of its resemblance to prefrontal cortex, the NCL has been the focus of considerable interest (Gtürkün, 1997; Durstewitz et al., 1999; Riters et al., 1999). The new term nidopallium serves to retain the existing abbreviations for the NCL, which is an advantage, given the many publications on this region (Table 4).
The term ectostriatum, broken into its prefix and root word, means “outside the striatum.” Because what has been called the ectostriatum is outside the striatum, as we define the striatum here, the term ectostriatum is actually semantically appropriate and could have been retained without any erroneous denotation. Nonetheless, the term ectostriatum was linked to the set of incorrect names for the pallium in birds by the root word “-striatum” and could be misconstrued as being part of the striatum. For this reason, the name for the ectostriatum was changed to entopallium, which means “within (ento-) the pallium.” This new term also retains existing abbreviations for this region and its prefix sounds similar to that for the ectostriatum (Table 4, Fig. 8B).
Although the term nucleus basalis as it has been used in birds to refer to a sensory structure of the pallium does not possess any root words implying an association with the basal ganglia, the name used for this sensory cell group needed to be changed because the Forum had already reserved that same name for the avian homologue of the basal forebrain cholinergic cell field in mammals (Table 2). The structure that has been called nucleus basalis in birds is not located in the subpallium and is not a cholinergic cell group, but rather is a trigeminorecipient pallial sensory cell group (Witkovsky et al., 1973; Wild et al., 1985) and, in some species, also a general somatosensory recipient nucleus (Wild et al., 1997, 2001). It also receives an auditory input from the nucleus of the lateral lemniscus, as well as an input from the vestibular system ((Arends and Zeigler, 1986; Wild and Farabaugh, 1996; Wild et al., 2001). To prevent confusion with the cholinergic nucleus basalis of mammals, the Forum recommended that the old nucleus basalis of birds be renamed the nucleus basorostralis of the pallium, which describes its position in the telencephalon, with no abbreviation change (Table 4, Figs. 7D, ,8A8A).
The nidopallial region containing the primary auditory thalamorecipient zone was neither recognized to be a distinct region nor assigned a name in Karten and Hodos (1967). However, the experimental work of Karten (1968) established that this zone largely coincides with the cytoarchitectonically defined Field L of Rose (1914), and this name subsequently became entrenched in the literature on this region (Bonke et al., 1979a; Kelley and Nottebohm, 1979; Brauth et al., 1987; Brauth and McHale, 1988; Fortune and Margoliash, 1992; Wild et al., 1993; Vates et al., 1996). For this reason, the Forum concluded that it would be disruptive to change the name of this region to an actual name rather than a letter, given the many studies devoted to it; thus, the Forum recommended that the name Field L be retained (Table 4, Figs. 7D, 8C,D).
The region of the songbird brain termed the HVC (Table 4, Fig. 7D) was first recognized as part of the telencephalic song control circuit by Nottebohm et al. (1976). It was (erroneously) thought to occupy a caudal part of the hyperstriatum ventrale and was thus named the hyperstriatum ventrale pars caudale, abbreviated HVc. Subsequent work, however, recognized that this region is in actuality located within the pallial field that has been called the neostriatum (Paton et al., 1981). To retain the abbreviation, which had already become entrenched, but to eliminate the inaccurate name, Nottebohm (1987) suggested calling this region the higher vocal center and abbreviating it with all capital letters. Subsequently, the concern was raised that the HVC was arguably not the apex of a hierarchy of vocal centers of the brain, making the name unwarranted (Margoliash et al., 1994). These issues are discussed by Fortune and Margoliash (1995) and by Brenowitz et al. (1997), who recommended use of “HVc” as the proper name for the nucleus. Since that time, however, some investigators have continued to use the term “high (or higher) vocal center” in published reports, whereas others have used “HVc” or “HVC” as a proper name. For the sake of consistency, the Forum suggested using HVC as the proper name and recommends against using HVc or any form of the term “higher vocal center.”
Several additional nuclei have been identified in songbirds that are important for auditory processing or are related to song learning and control, including the lateral magnocellular nucleus of the anterior neostriatum (lMAN or LMAN; Fig. 7D), the medial magnocellular nucleus of the anterior neostriatum (mMAN or MMAN), and the caudomedial neostriatum (NCM; Nottebohm et al., 1976; Bottjer et al., 1989; Mello et al., 1992; Mello and Clayton, 1994; Margoliash, 1997; Nordeen, 1997; Jarvis et al., 2000). With the renaming of neostriatum, the name for each of these is altered by substituting nidopallium for neostriatum, and the established abbreviation remains the same (Table 4). The nucleus interface (NIf) represents an additional major telencephalic constituent of auditory and song-control circuitry (Nottebohm et al., 1982; Vates et al., 1996). Although this nucleus, too, is located in what has been called the neostriatum, the word neostriatum does not appear in the outdated name for NIf. To emphasize its location, the Forum adopted for this structure the official name nucleus interface of the nidopallium (or its Latin equivalent), and thus its abbreviation does not need to be revised. For these abbreviations, as for others, the Forum recommends using capital letters for all main words represented in the abbreviations, with lower case only for subordinate words or letters. In the instance of the lateral magnocellular nucleus of the anterior nidopallium, this practice eliminates the confusion caused by the resemblance among the capital letter I, the lower case letter L, and the number 1.
Several nuclei have been identified in parrots that are devoted to vocal control, including the oval nucleus of the anterior neostriatum (NAo), the medial division of the oval nucleus of the anterior neostriatum (NAom), and the central nucleus of the lateral neostriatum (NLc; Striedter, 1994; Durand et al., 1997; Jarvis and Mello, 2000). With the renaming of the neostriatum, the name for each of these is altered by replacing neostriatum with nidopallium, and the established abbreviations remain the same, but with full capitalization.
It has been suggested that the archistriatum of the telencephalon in birds is at least partly comparable to the mammalian amygdala (Edinger et al., 1903; Edinger, 1908; Ariëns-Kappers, 1922; Ariëns-Kappers et al., 1936; Zeier and Karten, 1971; Bruce and Neary, 1995; Puelles et al., 2000), a structure that itself possesses both pallial and subpallial parts (Swanson and Petrovich, 1998; Puelles et al., 2000). Based on neurochemical and developmental data, it seemed overwhelmingly clear to the Forum that all parts of the archistriatum (i.e., the regions with the word archistriatum in their name) as defined in Karten and Hodos (1967) and in Kuenzel and Masson (1988) are pallial (Puelles et al., 2000; Wada et al., 2001, Reiner et al., 2002; Sun et al., 2003). The Forum further concluded that the taenia (also called the nucleus taeniae) has typically been regarded as a part of the archistriatal complex, although this was not reflected in its name (Ariëns-Kappers et al., 1936; Zeier and Karten, 1971; Thompson et al., 1998; Chen et al., 1999; Absil et al., 2002) and that much or all of the taenia is subpallial (Foidart et al., 1999; Cobos et al., 2001b; Absil et al., 2002).
The Forum also concluded that hodological, developmental, neurochemical, and behavioral evidence supported the amygdaloid nature of the taenia and posterior archistriatum (Zeier and Karten, 1971; Veenman et al., 1995b; Lanuza et al., 2000; Puelles et al., 2000; Absil et al., 2002; Roberts et al., 2002). By contrast, the anterior, intermediate, and at least parts of the medial archistriatum have largely somatic features, making them unlike the amygdala in mammals (Zeier and Karten, 1971; Veenman et al., 1995b; Davies et al, 1997; Mello et al., 1998b; Reiner et al., 2002; Wada et al., 2001). Although it was acknowledged that perhaps the anterior, intermediate, and medial archistriatum were, nonetheless, homologous to some parts of the mammalian amygdala and that the differences between them and mammalian amygdala were the result of divergent evolution, the Forum acknowledged that this had not been demonstrated. In addition, even if such an evolutionary relationship was proved, the concern was expressed that it would be misleading and inappropriate to attach a name with visceral and limbic functional implications (i.e., the term amygdala) to a field with prominent somatic traits (Wild, 1993; Knudsen and Knudsen, 1996; Margoliash, 1997).
The Forum resolved these conflicting concerns to its satisfaction by accepting only the posterior part of the archistriatum as warranting the designation of amygdala and appending amygdala to the name of the taenia. For the remaining parts of the archistriatum, the Forum decided to replace archistriatum with the term arcopallium, with the prefix “arco-” referring to the arched contour of the upper boundary of the field. This choice does not foreclose the option of replacing “arco-” with “amygdalo-” for specific arcopallial subdivisions if evidence for such homology becomes convincing. The subpallial region inferior to what we now recommend be called the globus pallidus in birds was renamed the subpallial amygdala. Further comments about the name changes for individual parts of this arcopallial-amygdaloid field follow (Table 5, Figs. 5--99).
Karten and Hodos (1967) and Kuenzel and Masson (1988) both recognize an anterior archistriatum and show it as an anterior continuation of the intermediate archistriatum, thus questioning whether it is a distinct entity. The Forum saw this issue as separate from the simple matter of renaming this region as the anterior arcopallium (Table 5, Figs. 6B, ,7E7E).
This region was called the dorsal archistriatum in Karten and Hodos (1967), but in their reformulation of the archistriatum, Zeier and Karten (1971) renamed this region the dorsal part of the intermediate archistriatum. By neurochemistry and connections, it is a distinct region (Wächtler, 1985; Wächtler and Ebinger, 1989; Medina and Reiner, 1994; Reiner et al., 1994; Veenman et al., 1995b; Kröner and Güntürkün, 1999; Sun and Reiner, 2000). The Forum recommended this region be renamed the dorsal arcopallium (Table 5, Figs. 6F, ,7C,7C, ,8C,8C, 9A,B), without intermediate in the name, for the sake of simplicity and in recognition of its distinctness.
This region constituted the dorsal part of the ventral archistriatum in Karten and Hodos (1967), but in their reformulation of the archistriatum, Zeier and Karten (1967) renamed this region the intermediate archistriatum. By neurochemistry and connections, it is distinct from what has until now been called the dorsal intermediate archistriatum (Wächtler, 1985; Wächtler and Ebinger, 1989; Reiner et al., 1994; Veenman et al., 1995b; Kröner and Güntürkün, 1999; Sun and Reiner, 2000). The Forum recommended that the intermediate archistriatum be renamed the intermediate arcopallium, with “pars ventralis” not needed due to the deletion of “intermediate” from the name for the dorsal arcopallium (Table 5, Figs. 6F, ,7C,7C, ,8C,8C, 9A,B,D).
This region was largely subsumed within the more medial part of the ventral archistriatum in Karten and Hodos (1967), but in their reformulation of the archistriatum, Zeier and Karten (1971) renamed this region the medial archistriatum; they regarded it as visceral and limbic in its connections and thus amygdaloid in its nature. More recent studies, however, have suggested that at least part of the medial archistriatum may be somatic in its projections (Davies et al., 1997). Thus, it is unresolved whether the structure known as the medial archistriatum contains any subregions that are amygdaloid in their connectivity and neurochemistry, and certainly the boundaries between any putative limbic and somatic parts are unresolved. The Forum recommended, therefore, that a cautious approach be taken in renaming the medial archistriatum. We recommend calling it the medial arcopallium, until such time as its possible amygdaloid subdivisions are more clearly defined (Table 5, Figs. 6C, ,7F,7F, ,8C,8C, ,9B).9B). Note that the abbreviation recommended for the medial arcopallium is the same as that employed in the Karten and Hodos (1967) atlas for the anterior medial hypothalamic nucleus. We thus suggest that the latter nucleus be abbreviated AMH to avoid a conflict.
Specialized nuclei involved in vocalization or audition have been identified within the archistriatum intermedium of songbirds, parrots, hummingbirds, and pigeons (Jarvis et al., 2000; Wild et al., 1993). These include the robust nucleus of the archistriatum (RA) of male songbirds and females of some songbird species (Figs 7D, ,9D;9D; Nottebohm et al., 1976; Bottjer et al., 1989; Wild, 1993; Margoliash, 1997; Nordeen, 1997), the central nucleus of the anterior archistriatum in parrots (Aac; Paton et al., 1981; Striedter, 1994; Durand et al., 1997), the vocal nucleus of the archistriatum (VA) in hummingbirds (Jarvis et al., 2000), and the ventromedial and dorsomedial nuclei of the intermediate archistriatum (Aivm and Aidm) in pigeons (Wild et al., 1993). With the renaming of the archistriatum, the name for each of these is altered by replacing archistriatum with arcopallium, and the abbreviation remains the same, except for full capitalization (Table 5). Note that the archistriatal cup region adjacent to the RA has been recognized as part of the auditory pathway in songbirds and is equivalent to Aivm/Aidm in the pigeon (Kelley and Nottebohm, 1979; Mello et al., 1998b).
This region was included as part of the ventral archistriatum in Karten and Hodos (1967), but in their reformulation of the archistriatum, Zeier and Karten (1971) renamed this region the posterior archistriatum and regarded it as visceral and limbic in its connections, and thus amygdaloid in its characteristics. This view has been amply supported in subsequent studies (Berk and Hawkin, 1985; Veenman et al., 1995b; Davies et al., 1997; Dubbeldam et al., 1997; Lanuza et al., 2000; Roberts et al., 2002). For this reason, the Forum recommended that the posterior archistriatum be renamed the posterior pallial amygdala, with pallial in the name to identify it as part of the pallial amygdala in birds (Table 5, Figs. 8C,D, ,9C).9C). For this and other amygdaloid nuclei, the Forum assigned three-letter abbreviations to distinguish them from arcopallial nuclei.
The nucleus taeniae was not extensively discussed in Zeier and Karten (1971) and was thereby seemingly treated as separate from the archistriatal complex. The Forum recognized that any such implication was unintentional and that the nucleus taeniae has been typically regarded as part of the archistriatal complex by investigators of this brain region (Ariëns-Kappers et al., 1936). Thus, the nucleus taeniae is deserving of consideration in the renaming of the parts of the archistriatal complex. Of note, amygdaloid features of the nucleus taeniae have been shown by recent neurochemical, connectional, developmental. and functional studies (Balthazart et al., 1992, Balthazart et al., 1998; Cheng et al., 1999; Foidart et al., 1999; Lanuza et al., 2000; Cobos et al., 2001b, Absil et al., 2002; Roberts et al., 2002). For this reason, the Forum recommended that the nucleus taeniae be renamed the nucleus taeniae of the amygdala (Table 5, Fig. 6C, ,7F,7F, ,8C,8C, 9A,B,D). The Forum further acknowledges evidence that all or most of the medial part of the nucleus taeniae is subpallial in nature (Absil et al., 2001; Cobos et al., 2001b; Roberts et al., 2002) and therefore perhaps akin to part of the mammalian medial amygdala (a subpallial amygdalar subdivision).
A subpallial region ventral to the globus pallidus at the level of the occipitomesencephalic tract possesses the neurochemistry and connectivity of the extended amygdala of mammals, notably its sublenticular part (Berk, 1987; Wild et al., 1990; Molnár et al., 1994; Reiner et al., 1994; Atoji et al., 1996; Lanuza et al., 2000; Wada et al., 2001; Roberts et al., 2002). This region had been depicted as a medial part of the medial archistriatum in Figure 8 of Zeier and Karten (1971) but is shown as a subpallidal part of the striatum (specifically PA) in the Karten and Hodos atlas (1967). Whether birds possess a specific homologue of the mammalian central nucleus of the amygdala, a major subpallial amygdalar region, is yet unresolved (Wild et al., 1990), but the location of the SpA is more comparable to that of the sublenticular extended amygdala than to that of the central amygdaloid nucleus. Nonetheless, the evidence is strong that the aforesaid region below the globus pallidus in birds is both striatal and amygdaloid. For this reason, the Forum recommended that it be recognized as such and for now given the generic name of subpallial amygdala (Table 5, Figs 5F, ,6C,6C, ,7E,7E, ,9A).9A). Thus, the currently recognized amygdaloid complex in birds consists of the posterior pallial amygdala, the taenia amygdalae, and the subpallial amygdala.
There were no compelling reasons to revise the boundaries or nomenclature for the hippocampal complex in birds (Erichsen et al., 1991; Krebs et al., 1991; Szekely, 1999). There was also no evident reason to rename the piriform cortex, temporo-parieto-occipital area (TPO), or dorsolateral corticoid area (CDL; Table 6, Figs. 6C, 8C,D). The anterior-posterior extent of the piriform cortex was, however, recognized to be greater than shown in the various bird brain atlases (Reiner and Karten, 1985), and one recent study has suggested that the extent of the CDL is greater than is shown in standard avian brain atlases (Redies et al., 2001).
The Forum recognized that the new nomenclature for cell groups of the telencephalon in birds resulted in the need to change the names of some of the cell-free laminae between major telencephalic regions, and in other cases name changes might be welcome if they helped make the location and nature of the given lamina clearer. The same pertains to fiber tracts. The Forum recommended the following actions (Table 7).
The lamina separating what is now known to be the basal ganglia from what is now known to be the overlying pallium has been called the lamina medullaris dorsalis Karten and Hodos, 1967; Kuenzel and Masson, 1988). Because this term is cryptic in meaning, the Forum recommended a simple and informative new name, the lamina pallio-subpallialis or its English variant, the pallial-subpallial lamina (Table 7, Figs. 6A–D, 8A–C).
Because the dorsal archistriatum has been renamed the dorsal arcopallium, it was necessary to rename the cell-free border defining the upper edge of what is now called the arcopallium, which separates the arcopallium from what is now called the nidopallium (Zeier and Karten, 1971; Kuenzel and Masson, 1988). The new name, lamina arcopallialis dorsalis, retains the existing abbreviation (Table 7, Fig. 8C,D).
In changing the name of the hyperstriatum ventrale to the mesopallium, it became inappropriate to refer to the lamina between it and the structure now called the nidopallium as the lamina hyperstriatica. As a new term for this lamina, the Forum recommended the mesopallial lamina or, in Latin, lamina mesopallialis (Table 7, Fig. 8A–C). In this instance, an “a” has been added to the abbreviation to prevent confusion with the medial lemniscus and the nucleus lentiformis mesencephali, which have been abbreviated LM in atlases and studies of avian brains (Karten and Hodos, 1967).
Although it was possible to rename this lamina, which separates the mesopallium from the hyperpallium densocellulare (as we suggest they now be called), the name is not inconsistent with the new terminology and it is entrenched. For this reason and in the absence of consensus as to a new name that is more informative, the Forum recommended retaining this term (Table 7, Fig. 8A,B).
Although it was also possible to rename this lamina, which separates the hyperpallium densocellulare from the hyperpallium intercalatum (as we propose they now be called), the name is not inconsistent with the new terminology and it is entrenched. For this reason and in the absence of consensus as to a new name that was more informative, the Forum recommended retaining this term as well (Table 7, Fig. 8A,B).
The fiber-rich zone interposed between the entopallium and the lateral striatum primarily contains axons that had coursed from the nucleus rotundus of the thalamus to the telencephalon via the lateral forebrain bundle (or fasciculus prosencephali lateralis), in their passage to their termination in the structure that we recommend now be called the entopallium (Karten and Hodos, 1970; Nixdorf and Bischof, 1982). This fiber-rich zone has been termed the tractus thalamostriaticus (TTS). This name is inappropriate because the tract, in fact, courses to a pallial and not a striatal target (Karten and Hodos, 1970). The tract does not exclusively terminate within the entopallium, however, because it also appears to contain some fibers arising from the dorsolateral posterior thalamic nucleus (Gamlin and Cohen, 1986) that project to the nidopallial region just medial to the entopallium and some fibers from the principal optic nucleus of the thalamus that project to the Wulst (Karten et al., 1973). Thus, the Forum recommended that the name for this fiber bundle be changed to the tractus thalamopallialis, resulting in some change to the abbreviation for this tract (Table 7, Figs. 6A, ,8B8B).
This fiber bundle interconnects the dorsal and posterior pallium with the structure now termed the arcopallium. Hence, the name for this tract was changed by the Forum, substituting arcopallialis for archistriatalis and modifying the placement of the adjective in the Latin name slightly, so the existing abbreviation is retained for both the English and Latin forms of the new name (Table 7, Fig. 8C,D).
This fiber bundle interconnects the rostral pallium with the structure now termed the arcopallium. Hence, the name for this tract was changed by the Forum, substituting arcopallialis for archistriatalis, with the existing abbreviation thereby retained (Table 7, Figs. 6A,B, ,8B,8B, ,9D9D).
Pathway tracing studies have shown that the fiber tract called the TSM in actuality conveys axons from cells in the Wulst and hippocampus to brainstem sites, via axons that traverse the septum (Karten et al., 1973; Krayniak and Siegel, 1978a,b; Reiner and Karten, 1983; Wild and Williams, 2000). The Forum also observed that the “-mesencephalicus” part of the name may be somewhat misleading if taken literally, because the Wulst (primarily the HA) via the TSM also gives rise to projections to diencephalic sites and sites posterior to midbrain, including the upper spinal cord in at least some species (Karten et al., 1973; Reiner and Karten, 1983; Wild and Williams, 2000). Nonetheless, for continuity with the previous literature, the Forum concluded that it would be appropriate to rename the tractus septomesencephalicus as the tractus septopallio-mesencephalicus, with no change in abbreviation, to reflect its origin in the pallium (Table 7, Figs. 6B, 9B,C). Note that the portion of the TSM arising from the primary sensory/motor field within the Wulst that courses to upper spinal levels resembles the pyramidal tract of mammals (Medina and Reiner, 2000; Wild and Williams, 2000).
No revision was recommended by the Forum for the tractus occipitomesencephalicus, the fiber bundle conveying the descending projection of the arcopallium (Zeier and Karten, 1971), because the name is largely appropriate and it is well entrenched (Table 7, Figs. 6C, ,7E,7E, ,9A).9A). As in the case of the fiber bundle previously called the tractus septo-mesencephalicus, the Forum noted that the “-mesencephalicus” part of the name might be construed to suggest that the OM projects only to the midbrain, which is not the case (Zeier and Karten, 1971; Wild and Farabaugh, 1996). Nonetheless, the value in retaining the abbreviation for OM was thought to outweigh any slightly misleading connotation in the term OM. In addition, the Forum took the view that the term occipito-mesencephalic tract should not be taken to mean that it interconnects the occipital pole of the telencephalon only with the midbrain, only that the tract extends from telencephalic to midbrain levels.
No revision was recommended by the Forum for the tractus occipito-mesencephalicus, pars hypothalami, the fiber bundle conveying the descending projection of the arcopallium and amygdala to the hypothalamus (Zeier and Karten, 1971), because the name is largely appropriate and it is well entrenched (Table 7). In addition, the Forum took the view, as noted in Zeier and Karten (1971), that some part of the HOM corresponds to the stria terminalis of mammals. At this time, however, the precise limits of the stria terminalis in birds are uncertain, and the Forum concluded that it was premature to identify a specific fiber tract in birds as the stria terminalis.
Our understanding of avian brain organization and function has advanced enormously in the past 100 years ((Edinger et al., 1903; Edinger, 1908; Ariëns-Kappers, 1922, 1928; Huber and Crosby, 1929; Ariëns-Kappers et al., 1936; Karten, 1969, 1991; Reiner et al., 1998a; Puelles et al., 2000). At the beginning of the 20th century, the intellectual abilities of birds were commonly held in low regard, and they were believed to be limited to an instinctive, inflexible behavioral repertoire (Edinger, 1903, 1908; Ariëns-Kappers, 1922, 1928; Huber and Crosby, 1929; Arïns-Kappers et al., 1936; Herrick, 1948, 1956). The telencephalon of birds was thought to be devoid of the neural equipment (i.e., cerebral cortex) that allows mammals to adapt their behaviors to their environments, and birds were regarded as instead possessing a telencephalon that consisted of an enormously hypertrophied basal ganglia. Because the basal ganglia were thought to be involved in control of unlearned, instinctual behaviors (Reiner et al., 1984a), the putative basal ganglia hypertrophy in birds appeared to account for what was taken to be their elaborate but stereotyped behavioral repertoire.
It is now evident that birds are not uniformly impoverished in their adaptive learning skills. Songbirds, parrots, and hummingbirds show vocal learning abilities not paralleled by any mammals other than humans and cetaceans (Farbaugh et al., 1994; Durand et al., 1997; Doupe and Kuhl, 1999; Pepperberg, 1999; Hile et al., 2000; Jarvis et al., 2000). Crows, a type of songbird, show the ability to make and use tools (Hunt, 1996, 2000; Weir et al., 2002), and parrots are capable of learning and communicating with simple semantic human language and show cognitive skills evident only in apes and cetaceans among nonhuman species (Pepperberg, 1999, 2002). In parallel with the growing awareness of avian behavioral sophistication has come growth in our understanding of the neural substrates within avian brains allowing such behavior. It has become clear that the telencephalon in birds does not consist of a hypertrophied basal ganglia, but rather possesses a well-developed pallial region that enables birds to perform remarkably complex behaviors. The pallial region in birds, however, does not have a layered organization, as it does in mammals (Karten, 1969, 1991; Reiner et al., 1984a), rather, in birds, the pallium is organized into a largely continuous field of nuclei. Although these nuclei appear to mediate the same functions as the cerebral cortex, they have a cytological appearance more like that of the basal ganglia, explaining the mistake of many earlier comparative neuroanatomists.
The errors in the 20th century terminology for the telencephalon and many brainstem cell groups in birds have perpetuated misconceptions about birds and avian brains and have obscured major conserved features of vertebrate brain evolution. The Avian Brain Nomenclature Forum was the culmination of growing awareness of these errors and their adverse impact on communication among scientists. The Forum sought to devise a new terminology that is free of errors and that promotes accurate understanding of avian brain organization and evolution. We have been scrupulous in our renaming to use only names implying homology that we are certain would not later themselves prove to be in error. We believe that the nomenclature changes we have devised can serve the field well, and we urge all investigators to use this new terminology. Further information and avian brain images depicting this new nomenclature are available on the Avian Brain Nomenclature Exchange website: http://avianbrain.org.
In changing the names of various structures of the forebrain, midbrain, and hindbrain, the Forum does not mean to suggest that these are the only names that would profit from change, only that these are the ones in most obvious need of immediate change. Noteworthy examples of other regions that merit scrutiny for possible revisions are the preoptic area and hypothalamus. Nonetheless, the Forum decided not to consider renaming structures in these regions because detailed descriptions of the cytoarchitecture, neurochemistry, and development of cell groups making up these regions are lacking for birds. Therefore, no recommendations are made at present about these brain areas. However, there are important questions to be resolved for birds, such as the location of the ventromedial hypothalamic nucleus, the subdivisions of the paraventricular nucleus, and the identity of the suprachiasmatic nucleus. Interested readers are directed to the following publications for a discussion of these issues: Berk and Butler (1981), Kuenzel and Van Tienhoven (1982), Berk and Finkelstein (1983), Mikami (1986), Cassone and Moore (1987), Norgren and Silver (1990), Balthazart et al. (1996), Panzica et al. (1996), and Abraham et al. (2002).
Finally, in revising the terminology, we do not mean either to discredit or to disrespect those neuroscientists before us who proposed and used this terminology, and we disavow any discourtesy to our forerunners. Those who preceded us in the study of avian brains, and developed the terminology that has been used for 100 years, were the leading comparative neuroanatomists of their time. If we today have been able to understand avian brain organization more accurately, it is because we have tools that were not available to them. Moreover, the work they did and the ideas they proposed laid an essential foundation, upon which we and others have been able to build in shaping the current view of avian brain structure and function. Thus, in closing this description of the new terminology for avian telencephalic and related brainstem areas, we offer tribute and thanks to Edinger, Rose, Ariëns-Kappers, Crosby, Huber, Kuhlenbeck, Källén, Nauta, Karten, and Hodos for their trailblazing and inspirational work on avian brains and their evolutionary relationship to mammalian brains (Fig. 10).
We thank the following individuals at Duke University for administrative, technical, and logistical support for the Forum: Deepa Bharanidharan (Jarvis Laboratory Associate in Research), Eunice Chang (Graduate Student), Margaret Couvillon (Graduate Student), Haruhito Horita (Graduate Student), Susan Havrilesky (Neurobiology Department Manager), Michael McElroy (Jarvis Laboratory Research Technician), Dawn Kernagis (Jarvis Laboratory Associate in Research), Lisa Moore (Jarvis Laboratory Manager), Martha Musson (Neurobiology Department Secretary), and the Netfriends computer assistants Ann Sink (Neurobiology Department coordinator), David Stokes (Web designer), and Tony Zimmermann (Jarvis Laboratory Research Analyst). We note the valuable contributions of Dr. Luis Puelles to the discussions on avian brain organization, development, and evolution that preceded the nomenclature meeting, and we thank Drs. Steve Brauth and Todd Roberts for making their data on the parrot telencephalon available to us prior to publication. A number of other researchers, too numerous to list here, made valuable contributions to the discussions in the years leading up to the Forum. Finally, we thank the Archives of the University of Alabama at Birmingham and the MIT Museum for the images of Elizabeth Crosby and Walle Nauta, respectively, in Figure 10. Preparation for the Forum, the Forum itself, and the dissemination of the conclusions of the Forum were supported by grants from NSF and NIH.
DEDICATION Finally, we dedicate this paper to Dr. Harvey J. Karten, in many respects the father of modern avian neuroanatomy. He has contributed vastly to the field and been a source of inspiration for many of us. We thank him for this and for his contributions to the nomenclature revision, and we hope our own efforts in this latter regard show him proper recognition and gratitude.
Grant sponsor: National Science Foundation; Grant number: IBN-0110894; Grant sponsor: National Institutes of Health; Grant number: 1R13-MH-64400-01.
Published online in Wiley InterScience (www.interscience.wiley.com).