Department of Genetics, University of Georgia, Athens, Georgia 30602 USA Nuclear genes encode two major forms of malate dehydrogenase (MDH; E.C. 1.1.1.37) in birds: S-MDH (or MDH-1), which is found in the soluble fraction of the cell cytoplasm; and M-MDH (or MDH-2), which is housed in mitochondria (Karig and Wilson 1971). Under standard starch-gel electrophoretic condi- tions, S-MDH typically migrates toward the anode while M-MDH migrates cathodally. In terms of gen- eral zymogram appearance and electromorph fre- quencies across avian taxa, both S-MDH and M-MDH are very conservative evolutionarily (Kitto and Wil- son 1966, Aquadro and Avise 1982, Kuroda et al. 1982). In particular, S-MDH assayed in pH 7.0 gels exhib- ited the same common electromorph in representa- tives of 22 avian orders (Kitto and Wilson 1966). Al- though additional S-MDH variation was detected under the multiple electrophoretic conditions em- ployed by Aquadro and Avise (1982), S-MDH re- mains extremely conservative in comparison with most other enzyme systems. Kitto and Wilson (1966) first noticed an unusual S-MDH zymogram pattern in three species of Pici- formes. Unlike the typical avian S-MDH pattern, which consists of an intense major band with one or two much lighter anodic subbands, the Piciformes were "exceptional in having two S-MDH subbands  Present address: Section of Genetics and Devel- opment, Cornell University, Ithaca, New York 14853 USA. of approximately equal intensity" (Kitto and Wilson 1966). This distinctive S-MDH pattern was subse- quently observed in a Japanese woodpecker (Kuroda et al. 1982: fig. 3) and in three species of North Amer- ican woodpeckers (Aquadro and Avise 1982). Phylogenetic relationships of the woodpeckers and allies have been the subject of intense debate (Simp- son and Cracraft 1981, Swierczewski and Raikow 1981, Olson 1983, Raikow and Cracraft 1983, Sibley and Ahlquist 1986). In most traditional taxonomic lists, six families are included in the Piciformes: Galbuli- dae (jacamars), Bucconidae (puffbirds), Capitonidae (barbers), Indicatoridae (honeyguides), Ramphasti- dae (toucans), and Picidae (woodpeckers). Recently, this classification has been challenged on the grounds that the assemblage may be polyphyletic (Olson 1983, Sibley and Ahlquist 1986). Raikow and Cracraft (1983), however, defended the validity of the synapomorphs (hypothesized shared-derived traits) used to infer monophyly of the Piciformes. Additional questions concern the relationships of true piciform families to one another. In an effort to help resolve some of these phylogenetic issues, we surveyed members of all the taxa listed above, plus other possible piciform allies, for occurrence of the distinctive and obviously de- rived S-MDH woodpecker pattern. Frozen tissue samples (usually heart, liver, or mus- cle) were homogenized separately in an equivalent volume of grinding solution (0.01 M Tris, 0.001 M EDTA; pH 6.8) and centrifuged at 49,000 g. Electro- r+ Fig. 1. Representative S-MDH zymogram pat- terns in various species of "piciform" birds. From left to right: (1) Melanerpes carolinus (Picidae); (2) Galbula ruficauda (Galbulidae); (3) Veniliornis nigriceps (Pici- dae); (4) Melanerpes carolinus (Picidae); (5) Melanerpes cruentatus (Picidae); (6) Piculus rivolii (Picidae); (7) and (8) Chelidoptera tenebrosa (Bucconidae); (9) Monasa ni- grifrons (Bucconidae); (10) Melanerpes carolinus (Pici- dae); (11) Andigena cucullata (Ramphastidae); (12) Pteroglossus castanotis (Ramphastidae); (13) Ramphastos culminatus (Ramphastidae); (14) Malacoptila semicincta (Bucconidae). The 3-band S-MDH pattern is evident in lanes 1, 3-6, and 10-13. The typical S-MDH pat- tern in nonpiciform birds is pictured in fig. 1 of Aquadro and Avise (1982) and is like that in lanes 2, 7-9, and 14 above. phoresis of supernatant extracts involved horizontal starch gels (12.5%), run at 75 mA for 5-7 h. The fol- lowing amine-citrate (A-C) buffer system was em- ployed routinely: electrode buffer, 0.04 M citric acid, pH adjusted to 6.1 with N-(3-aminopropyl) morpho- line (purchased from Aldrich Chemicals); gel buffer, 1 to 19 dilution of electrode buffer. This system was chosen because, among 12 buffers utilized previ- ously, it provided the best resolution for S-MDH (Aquadro and Avise 1982). Nonetheless, other buffer conditions yielded comparable results. MDH was stained using the recipe of Selander et al. (1971). The avian material consisted of 29 species that rep- resented 22 of the 74 piciform genera (30%) recog- nized in most current checklists. In addition, mem- bers of Coliiformes and several families of Coraciiformes (possible relatives of the Piciformes) were surveyed (Table 1). The samples originated from three continents: Africa, North America, and South America (Table 1). S-MDH zymograms.--The distinctive "woodpeck- er-type" S-MDH electrophoresed with the amine-ci- trate buffer exhibits three major bands (Fig. 1), the two most anodal of which apparently correspond to the "two subband" zymogram pattern originally no- ticed by Kitto and Wilson (1966). This pattern is readily distinguishable from the usual S-MDH pat- tern in other birds, which on A-C gels consists of a single intense band, with occasional, much lighter surrounding bands observable upon heavy staining (Fig. I). In both the Downy Woodpecker (Picoides pubescens) and Northern Flicker (Colaptes auratus), we assayed Fig. 2. Two (among several) possible phylogenies for "piciform" birds. (A) Inferred from a traditional classification proposed by Peters (1948). (B) Proposed by Simpson and Cracraft (1981) and independently by Swierczewski and Raikow (1981). The postulated origin of the derived 3-band S-MDH zymogram is indicated by the black rectangle. pectoral muscle, leg muscle, heart, liver, gizzard wall, intestine wall, eye, and brain. There were no ob- vious, differential tissue specificities for any S-MDH isozymes (although all bands were darkest in the muscle and heart samples). In the woodpecker-type S-MDH zymogram, all three bands were invariably present, and the middie band stained most intensely. Thus, the gel pattern gives the impression of "fixed heterozygosity" for S-MDH, which is known to be a dimeric molecule in birds. One plausible explanation is that woodpeckers possess a gene duplication for S-MDH, such that the middle band in the zymogram represents an inter- locus heterodimer, flanked by the two respective in- tralocus homodimeric products. Similar zymogram patterns and assembly of interlocus hybrid molecules are known in some fishes that carry an S-MDH gene duplication (Bailey et al. 1970, Wheat et al. 1971). If the S-MDH gene is also duplicated in woodpeckers, individuals detectably heterozygous at either locus should exhibit a 6-band zymogram (three homo- and three heterodimers). Unfortunately, despite assay of more than 100 piciform specimens (Table 1), we did not detect any heterozygotes. This may not be sur- prising, however, because S-MDH is essentially monomorphic in most avian species (Aquadro and Avise 1982). T^BIE 1. Woodpeckers and possible nonpasseriform allies assayed for presence vs. absence of the unique S-MDH isozyme pattern. Unique "woodpecker" S-MDH Taxon n Continent of collection zymogram? Piciformes Picidae (woodpeckers) Colaptes auratus  3 Dryocopus pileatus 1 Melanerpes carolinus 34 Melanerpes erythrocephalus 7 Melanerpes formicivorus 11 Picoides pubescens 21 Picoides scalaris 5 Picoides villosus 5 Sphyrapicus varius 2 Melanerpes cruentatus 1 Piculus rivolii 1 Veniliornis nigriceps 1 Veniliornis passerinus 1 Campethera abingoni 1 Dendropicos fuscescens 1 Indicatoridae (honeyguides) Indicator minor 2 Capitonidae (barbers) Lybius torquatus b 1 Trachyphonus vaillantii 3 Tricholaema leucomelas 1 Rhamphastidae (toucans) Andigena cucullata 1 Pteroglossus castanotis 1 Ramphastos culminatus 2 Ramphastos toco  1 Bucconidae (puffbirds) Chelidoptera tenebrosa 2 Malacoptila semicincta 1 Monasa nigrifrons 1 Nonnula ruficapilla 1 Nystalus striolatus 1 Galbulidae (jacamars) Galbula ruficauda 1 Coraciiformes Alcedinidae (kingfishers) Ceryle alcyon 1 Bucerotidae (hornbills) Tockus erythrorhynchus 2 Tockus fiavirostris 1 Coraciidae (rollers) Coracias caudata 1 Momotidae (motmots) Momotus momota  ! Phoeniculidae (wood-hoopoes) Phoeniculus purpureus 2 North America North America North America North America North America North America North America North America North America South America South America South America South America Africa Africa Africa Africa Africa Africa South America South America South America South America South America South America South America South America South America South America North America Africa Africa Africa South America Africa Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes NO No No No No No No No No No No T^nLE 1. Continued. Unique "woodpecker" S-MDH Taxon n Continent of collection zymogram? Upupidae (hoopoe) Upupa epops  1 Africa No Coliiformes Coliidae (colies) Colius striatus 3 Africa No Urocolius indicus 1 Africa No One specimen also assayed by Kitto and Wilson (1966). Assayed by Kitto and Wilson (1966) but not in the present study. On the other hand, from available data we cannot exclude the possibility that some "epigenetic" phe- nomenon, such as posttranslational protein modifi- cation, might account for the multiplicity of S-MDH bands in woodpeckers. For example, subbands of al- cohol dehydrogenase in Drosophila result from the binding of a NAD-carbonyl complex to the protein (Everse et al. 1971). Regardless of its particular mo- lecular basis, the unique S-MDH zymogram pattern should provide an informative phylogenetic marker. S-MDH and piciform phylogeny.--The distinctive, 3-band S-MDH zymogram was present in all species of Picidae, Indicatoridae, Capitonidae, and Ramphas- tidae assayed, but not in the Bucconidae or Galbuli- dae (Table 1; see examples in Fig. 1). The "wood- pecker-like" S-MDH pattern was absent from all six families of Coraciiformes, and from the two repre- sentatives of the Coliiformes (Table 1). Together with taxa previously assayed (Table 2), S-MDH has now been surveyed in 26 of the approximately 27 living avian orders, and in 16 families of Passeriformes. Only in a subset of the Piciformes--woodpeckers, honey- guides, barbets, and toucans--has the distinctive 3-band S-MDH pattern been observed. One traditional and widely recognized classifica- tion of Piciformes (by Peters 1948; see also Wetmore 1960) implies the phylogeny shown in Fig. 2A. It is incompatible with the observed taxonomic distribu- tion of the 3-band S-MDH pattern, provided this MDH pattern has not been lost secondarily by the common ancestor of jacamars and puffbirds. The like- lihood of such secondary loss is difficult to assess critically. Note that the 3-band S-MDH pattern must be very old, however, because it is present in four families and in both Old and New World Piciformes. Sibley and Ahlquist (1985, 1986) estimated that some of these groups diverged nearly 80 million years ago. The earliest known fossils referable to the true Pici (as defined below) are about 50 million years old (S. L. Olson pers. comm.). In any event, the retention of the unusual S-MDH genotype over such time spans suggests that it may have some adaptive significance. If true, the probability of secondary loss of the pat- tern presumably would be low. Cladistic analyses of skeletal morphology (Simp- son and Cracraft 1981) and limb myology (Swier- czewski and Raikow 1981) of the Piciformes pro- duced the phylogeny shown in Fig. 2B. The observed distribution of the 3-band S-MDH pattern is sup- portive of the Pici clade composed of Ramphastidae, Capitonidae, Indicatoridae, and Picidae. Together with results from DNA-DNA hybridizations (Sibley T^nLE 2. Avian taxa (orders and passeriform fami- lies) previously assayed for S-MDH isozyme pat- terns. No assayed representatives of these groups exhibited the S-MDH pattern characteristic of woodpeckers. Tinamiformes a Passeriformes d Rheiformes a Eurylaimidae  Apterygiformes a Pipridae  Struthioniformes  Tyrannidae a Gaviiformes a,b Pittidae a Podicipediformes  Hirundinidae a,b Sphenisciformes  Corvidae a,b Procellariiformes a Paridae b Pelicaniformes  Troglodytidae b Ciconiiformes  Muscicapidae -,b Anseriformes a, Mimidae ,b Falconiformes ,b Bombycillidae b Galliformes a,b Sturnidae -, Gruiformes , Picathartidae  Charadriiformes  Nectariniidae  Columbiformes , Emberizidae ,b Psittaciformes  Passeridae , Cuculiformes , Strigiformes , Caprimulgiformes  Apodiformes  Trogoniformes   From Kitto and Wilson (1966),  From Aquadro and Arise (1982).  From Kuroda et al. (1982: fig. 3). a Olson (1983) suggested that the closest living relatives of the Pici (as in Fig. 2B) may be found among the Passeriformes. and Ahlquist 1986), and further analyses of mor- phology (Olson 1983), the monophyletic status of the Pici now seems well established. Most of the further debate about piciform phylogeny concerns whether the Galbulae and Pici are sister groups (as argued by Raikow and Cracraft 1983), or whether each has clos- est avian relatives elsewhere (as argued by Olson 1983, and Sibley and Ahlquist 1986). Wiley (1981) noted that "one true synapomorphy is enough to define a unique genealogical relation- ship. The problem is that the synapomorphies we hypothesize may or may not be true synapomor- phies." Much of the debate about piciform phylog- eny has centered on whether a few particular mor- phological features reflect shared ancestry or convergence. Sibley and Ahlquist (1983) argued that quantitative DNA-DNA hybridization values indi- cate true cladistic relationships because they are pro- portional to the relative times of divergence between taxa. They also predicted that qualitative morpholog- ical or genetic traits, properly interpreted, "will prove to be congruent with the DNA evidence in all cases" (Sibley and Ahlquist 1986). In this instance at least, the derived S-MDH genotype (as well as certain pre- viously studied morphological traits) are indeed con- gruent with the DNA-hybridization data in predict- ing monophyletic status for a Pici assemblage composed of the toucans, barbets, honeyguides, and woodpeckers. We thank Michael Braun and Herb Dessauer of the Louisiana State University (LSU) Museum of Frozen Tissues, and the following collectors of the South American birds: J. V. Reinsen, T. S. Schulenberg, D. A. Weidenfield, S. M. Lanyon, and A. P. Capparella. The LSU collections were made possible by cooper- ation of the Bolivian government and the Direccin de Ciencia y Tecnologla, through Gastn Bejerano, and the Academia Nacional de Ciencias, La Paz. We thank Alan Kemp and staff of the Transvaal Museum, Pretoria, South Africa. We also appreciate the efforts of Storrs Olson, Phil Angie, Nancy Joste, Peter Stan- gel, and Bobby Crawford, who provided specimens. Storrs Olson made many helpful suggestions for the manuscript. Work was supported by NSF grants BSR- 8217291 and BSR-8603775. LITERATURE CITED AQUADRO, C. F., & J. C. AVISE. 1982. Evolutionary genetics of birds. VI. A reexamination of protein divergence using varied electrophoretic condi- tions. Evolution 36: 1003-1019. BAILEY, G. S., A. C. WILSON, J. E. HALvER, & C. L. JOHNSON. 1970. Multiple forms of supernatant realate dehydrogenase in salmonid fishes. J. Biol. Chem. 245: 5927-5940. EVERSE, J., L. ZOLL, L. KAHAN, & N. KAPLAN. 1971. Addition products of diphosphopyridine nu- cleotides with substrates of pyridine nucleotide linked dehydrogenases. Bioorgan. Chem. 1: 207- 233. KARIG, L. M., & A. C. WILSON. 1971. Genetic varia- tion in supernatant realate dehydrogenase of birds and reptiles. Biochem. Genet. 5: 211-221. KITTO, G. B., & A. C. WILSON. 1966. Evolution of realate dehydrogenase in birds. Science 153: 1408-1410. KURODA, N., R. KAKIZAWA, H. HORI, Y. OSAKA, N. USUDA, r S. UTIDA. 1982. Evolution of mito- chondrial realate dehydrogenase in birds. J. Ya- mashina Inst. Ornithol. 14: 1-15. OLSON, S. 1983. Evidence for a polyphyletic origin of the Piciformes. Auk 100: 126-133. PETERS, J. J. 1948. Check-list of birds of the world, vol. 6. Cambridge, Massachusetts, Harvard Univ. Press. RAIKOW, R. J., & J. CRACRAFT. 1983. Monophyly of the Piciformes: a reply to Olson. Auk 100: 134- 138. SELANDER, R. K., M. H. SMITH, S. Y. YANG, W. E. JOHNSON, & J. B. GENTRY. 1971. Biochemical polymorphism and systematics in the genus Peromyscus. I. Variation in the old-field mouse (Peromyscus polionotus). Stud. Genet. VI, Univ. Texas Publ. 7103: 49-90. SIBLEY, C. G., & J. E. AHLQUIST. 1983. Phylogeny and classification of birds based on the data of DNA- DNA hybridization. Pp. 245-292 in Current or- nithology, vol. 1 (R. F. Johnston, Ed.). New York, Plenum. ß & --. 1985. The relationships of some groups of African birds, based on comparisons of the genetic material, DNA. Pp. 115-161 in Proc. Intern. Syrup. African Verts. (K.-L. Schuchmann, Ed.). Bonn, West Germany, Publ. A. Koenig Mus. , & --. 1986. Reconstructing bird phy- logeny by comparing DNA's. Sci. Amer. 254: 82- 93. SIMPSON, S. F., & J. CRACRAFT. 1981. The phyloge- netic relationships of the Piciformes (class Aves). Auk 98: 481-494. SwIERCZEWSKI, E. V., & R. J. RAKOW. 1981. Hind limb morphology, phylogeny, and classification of the Piciformes. Auk 98: 466-480. WETMORE, g. 1960. A classification for the birds of the world. Smithsonian Misc. Coil. No. 139. WHEAT, T. E., W. F. CHILDERS, E. T. MILLERß r G. S. WHITT. 1971. Genetic and in vitro molecular hybridization of realate dehydrogenase iso- zymes in interspecific bass (Micropterus) hybrids. Anita. Blood Grps. Biochem. Genet. 2: 3-14. WILEY, E. D. 1981. Phylogenetics. New York, John Wiley and Sons. Received 18 July 1986, accepted 26 September 1986.