Variation and hybridization in Green Heron (Butorides virescens) and Striated Heron (B. striata) in central Panama, with comments on species limits

Floyd E. Hayes • Department of Biology • Pacific Union College • Angwin, California 94508 • (Corresponding author; [email protected]) Douglas E. Weidemann • Department of Biology • Pacific Union College • Angwin, California 94508 CURRENT ADDRESS: Department of Biology • University of Miami • Coral Gables, Florida 33146 Dustin S. Baumbach • Department of Biology • Pacific Union College • Angwin, California 94508 CURRENT ADDRESS: Marine Research Group • Department of Earth and Biological Sciences • Loma Linda University • Loma Linda, California 92350 Richard D. Tkachuck • Cyndy M. Tkachuck • 1406 Clover Leaf Road, Locke, New York 13092

Abstract with a score of 5 comprised 14% of the thus supporting their current treatment as The rufous-necked Green Heron (Butorides population. The proportion of intermediate distinct species. virescens) of North America and the scores of 4–6 was slightly but not signifi- Caribbean hybridizes with the gray-necked cantly lower than a sample of 44 museum Introduction Striated Heron (B. striata) of South America specimens collected from 1908–1966 (33% Two species of Butorides herons are currently in central Panama and on islands of the versus 43%). The increased variability and recognized: the rufous-necked Green Heron southern Caribbean. Based on a color pho- intermediacy of neck color within the con- (B. virescens) of North America and the tograph of nine voucher specimens used as a tact zone strongly implies that hybridization West Indies, and the gray-necked Striated hybrid index, we obtained neck color scores still occurs between the two species in cen- Heron (B. striata) of South America (includ- (ranging from gray to purplish brown) of tral Panama. Because random mating tends ing dark B. s. sundevalli of the Galapagos 42 live individuals photographed in central to reduce variability around an intermediate Islands) and the Old World (A.O.U. 1998, Panama during 10-16 July 2011. Neck color phenotype, the current full range of phe- Banks et al. 2003). They were usually treated scores ranged from 1–8 (x=3.7, SD=2.2). notypes among herons presumably breed- as distinct species until Payne (1974) pro- Presumed B. striata with scores of 1–3 com- ing in central Panama suggests a tendency vided evidence of extensive hybridization prised 52% of the population, presumed B. toward assortative mating despite frequent where their ranges meet in central Panama, virescens with scores of 7–8 comprised 14% hybridization. The two taxa appear to have several southern Caribbean islands, and of the population, and presumed hybrids achieved essential reproductive isolation, coastal northern South America. Based on

Score: 1 Score: 2 Score: 3

2 north american birds variation & hybridization in Green Heron and Striated Heron

Payne’s (1974) conclusions, the American mens and live individuals in the field Neck color variation (counter-clockwise Ornithologists’ Union (1976, 1983) merged with a color photograph of Payne’s (1974) from bottom left of page 00, scores Green Heron and Striated Heron into a sin- voucher specimens used as a hybrid index. 1 through 8) of potentially breeding adult Butorides herons along a 6-km gle species, Green-backed Heron (B. striata). In Trinidad, the population is dominated section of the Chagres River in central Specimens from the Panamanian con- by B. striata; in nearby Tobago, the popula- Panama on 14 and 15 July 2011. tact zone were subsequently re-analyzed by tion is dominated by B. virescens, but with Photographs by Floyd Hayes. Monroe and Browning (1992), who con- an increase in variability and intermediacy cluded that Payne’s (1974) voucher speci- suggestive of hybridization. Hayes (2006) Score: 8 mens used as a hybrid index included juve- demonstrated a significantly lower pro- niles and did not represent a continuous se- portion of intermediate phenotypes in the ries. Monroe and Browning’s (1992) conclu- current Tobago population compared with sions that B. virescens and B. striata seldom museum specimens, suggesting a shift hybridized and should be regarded as dis- within the past century toward relatively tinct species were accepted by the American “pure” phenotypes. Hayes (2006) con- Ornithologists’ Union (1993, 1998), which cluded that the two taxa tended to mate as- re-split B. striata into Green Heron (B. vires- sortatively and therefore appeared to have cens) and Striated Heron (B. striata). achieved essential reproductive isolation, Hayes (2002) re-examined Payne’s (1974) supporting their current treatment as dis- voucher specimens and published a photo- tinct species. graph demonstrating that all had attained In this study we review the histori- adult neck coloration and appeared to rep- cal status of Butorides herons in Panama resent a continuous series. A re-analysis of and document current variation among Payne’s (1974) data demonstrated increased Butorides herons in the hybrid zone of cen- variability and intermediacy in the contact tral Panama. We compare our data with zone between B. virescens and B. striata, im- specimens collected >45 years ago to infer Score: 7 plying hybridization (see, e.g., Schueler and the extent of hybridization, gene flow, and Rising 1976). Hayes (2002) noted that phe- assortative mating between the two taxa notypically “pure” B. virescens and B. striata and to assess whether a recent shift in the coexisted within the contact zone, suggest- direction of gene flow has occurred in the ing that assortative mating occurred, but region, based on scoring neck coloration noted that the sample size of museum speci- and inferring gene flow from phenotype. mens was small and it remained uncertain whether both parental phenotypes actually Historical status in Panama bred within the hybrid zone. Early ornithologists were confused by At the eastern end of the hybrid zone, the status of Butorides herons in Panama. Hayes (2006) analyzed historic and cur- Thayer and Bangs (1906) identified a rent variability of the two taxa in Trinidad specimen collected in central Panama and Tobago by comparing museum speci- in 1904 as B. striata. Oberholser (1916) Score: 6 Score: 4 Score: 5

Volume 67 (2013) • Number 1 3 variation & hybridization in Green Heron and Striated Heron

described rufous-necked together,” but dismissed B. specimens from south- s. patens as individual varia- ern Central America and tion within B. s. striata. northern South America, Using a series of voucher including ten specimens specimens for scoring neck from Panama, as a new coloration on a scale of subspecies of B. virescens, 1–9 (from gray-necked B. which he named B. v. hyper- striata to rufous-necked B. notius. Oberholser (1916) virescens), Payne (1974) re- also described 22 rufous- ported considerable varia- necked specimens from the tion in neck color scores Pearl Islands in the Bay of among 44 specimens from Panama as a new endemic central Panama, collected subspecies of B. virescens, from 1908–1966 (dates which he named B. v. mar- from online specimen cata- garitophilus. Stone (1918) logs), with a mean score of reported two specimens of 4.8 (SD=2.1, range=1–9; B. striata and one specimen Table 1). A subset of 15 of B. v. hypernotius collected specimens collected from in central Panama in 1911 May–September, which and 1912. Hallinan (1924) probably eliminated mi- Figure 1. Sampling localities of Butorides herons in Panama during 10–16 July 2011. collected five specimens of Numbers represent the number of samples at each locality. gratory B. v. virescens, B. virescens and observed averaged slightly lower two active nests with eggs and chicks at Colorado Island and reported second-hand scores (x=4.1, SD=1.6, range=2–7; Table Balboa, central Panama, on 13 May 1916. accounts of “supposedly” B. virescens nests 1). Payne’s (1974) data revealed a cline of Griscom (1929) collected a specimen with eggs on 24 March (T. Gilliard) and 28 predominantly rufous-necked birds in west- identified asB. s. striata at El Real, Darién, April 1935 (Skutch). ern Panama to predominantly gray-necked in eastern Panama, which he claimed to be Wetmore (1965) summarized data from birds in eastern Panama, with a higher pro- the first report from Central America, but many new specimen and sight records of portion of intermediate phenotypes (scores overlooked earlier reports from Panama by four taxa in Panama: (1) migrant B. v. vi- of 4–6) in central Panama than in western Thayer and Bangs (1906) and Stone (1918). rescens from North America, a common Panama. Payne (1974) pointed out that B. Griscom (1929) believed Hallinan’s (1924) winter visitor from October to April; (2) v. margaritophilus from the Pearl Islands was specimens of B. virescens represented a new resident B. v. maculatus, common west of similar to populations from other Caribbean subspecies of B. striata, which he described the Canal Zone but ranging eastward to islands and later subsumed B. v. maculatus and named as B. s. patens. Griscom (1935) near Colombia; (3) resident B. v. margari- and B. v. margaritophilus into B. v. virescens later summarized the status of all Butorides tophilus, endemic to the Pearl Islands; and (Payne 1979). taxa in Panama. Hellmayr and Conover (4) resident B. s. striata, ranging from cen- Little has been published on the status of (1948) synonymized B. v. hypernotius with tral Panama eastward and more common Butorides herons in Panama during the past B. v. maculatus of the West Indies, attributed toward the east. Wetmore (1965) found a few decades. Recent field guides (Ridgely B. striata specimens collected by Thayer and nest of B. virescens with eggs in west-central 1976, Angehr and Dean 2010) tend to echo Bangs (1906) and Hallinan (1924) to B. s. Panama on 25 February 1956 and attrib- the conclusions of Wetmore (1965) and patens, and regarded Griscom’s (1929) speci- uted earlier reports of nesting B. virescens Payne (1974). Willis and Eisenmann (1979) men of B. s. striata as the only record from on Barro Colorado Island to B. s. striata. noted that B. virescens had become rare and Panama, but overlooked Stone’s (1918) re- Wetmore (1965) synonymized B. s. patens B. striata was nearly extirpated at Barro port. with B. s. striata and suggested the possi- Colorado Island in central Panama, due to Van Tyne (1950) obtained eight specimens bility of “occasional mixed mating among” competition for food with an introduced and “many satisfactory sight records” of B. B. virescens and B. striata “when they range species of peacock bass (Cichla pleiozona). striata on Barro Colorado Island in central Panama, Neck Color Scores reported a nest of B. striata with eggs on 28 July 1925 Time period 1 2 3 4 5 6 7 8 9 and half-grown fledglings on 1908–1966 (May–Sep only) 0 3 3 3 3 2 1 0 0 11 August 1925, stated that a B. virescens was not a breed- 1908–1966 (year-round) 2 4 6 11 6 2 6 6 1 ing resident, and questioned the validity of B. s. patens. 2011 (July) 7 10 5 3 6 5 5 1 0 Eisenmann (1952) reported both resident and migra- a : includes one specimen scored as “4–5” by Payne (1974) tory forms of B. virescens and Table 1. Frequency of neck color scores of Butorides herons in central Panama based on museum specimens resident B. striata at Barro collected during 1908–1966 (Payne 1974) and photographs of live birds in 2011 (this study).

4 north american birds v ariation & hybridization in Green Heron and Striated Heron

Methods Mann-Whitney U tests (z statistic) and two- m of each other; thus, none were suspected During 10–16 July 2011, we searched sample chi-square tests (x2 statistic) were of being mated pairs. One apparently nest- for Butorides herons in wetlands in cen- used to compare the distribution of neck col- ing individual flew with a stick in its bill, tral Panama between 78° 45’ and 80° 00’ oration scores between current and historical but we could not determine its identity. Most W longitude. Searches were conducted by time periods (Zar 1998). individuals were adults. We obtained pho- foot, car, and motorboat. A GPS unit was tographs of three juveniles (excluded from used to obtain coordinates of different sam- Results our sample) along the Chagres River near pling locations. For each Butorides found, We obtained neck color scores from 42 indi- Gamboa on 14 and 15 July. Two had mostly we attempted to obtain digital photographs viduals in central Panama, all on the east side attained adult neck coloration, which was clearly illustrating its neck coloration. We of the Panama Canal between 78° 54.42’ and bright rufous, clearly identifying them as B. tried to avoid sampling the same bird twice 79° 41.54’ W longitude (Fig. 1). All neck virescens. The third also had a rufous neck and scrutinized photographs of similar ap- color scores were obtained during 10–16 but with heavier streaking and was almost pearing birds to ensure that each sample July 2011, and most of our samples (79%) certainly B. virescens. was a unique individual (e.g., differences were obtained from a boat along a 6-km sec- in lore coloration, distribution of coloring tion of the Chagres River east of Gamboa D iscussion on neck, pale edges of wing coverts, etc.). (Figure 1). Our data revealed considerable Although some criticize the use of hybrid We compared each photographed indi- variability in neck color scores, ranging indices as too crude and subjective, inde- vidual with a photograph of a series of nine from 1–8 (x=3.7, SD=2.2; Frontispiece and pendent studies yield nearly identical re- voucher specimens used by Payne (1974) Table 1). Mean neck color scores were low- sults (Corbin and Barrowclough 1977). as a hybrid index in which neck coloration er, although not significantly, than those of Independent scoring of neck coloration of was scored on a scale of 1–9, ranging from museum specimens collected during May– museum specimens of Butorides herons from gray to dark purplish-brown (see Figure 1 of September (x=4.1; z=0.72, P=0.47). Mean Panama using voucher specimens (Monroe Hayes 2002). Because juveniles and imma- neck color scores were significantly lower and Browning 1992), and from Trinidad and tures have streaked necks (always browner than museum specimens collected year- Tobago using a color photograph of Payne’s than adults of S. striata), only adults and round (x=4.8; z=2.29, P=0.02), presumably (1974) voucher specimens (Hayes 2006), subadults that had fully acquired adult neck reflecting a lower proportion (or absence) of yielded results very similar to those of Payne coloration (Hayes 2002) were scored. When migratory B. v. virescens from North America (1974). Independent scoring of neck color- an individual appeared intermediate in neck in our sample. Presumed B. virescens with ation of live Butorides herons in the field in coloration between two voucher specimens, scores of 7–8 comprised 14% of the popula- the United States Virgin Islands (Hayes and we chose the specimen it most closely re- tion, presumed B. striata with scores of 1–3 Hayes 2006) and in Trinidad and Tobago sembled. Our neck color scores were based comprised 52% of the population, and pre- (Hayes 2006) using a color photograph on a group consensus by the authors. sumed hybrids with a score of 5 comprised of Payne’s (1974) voucher specimens also According to Payne (1974), specimens 14% of the population. No birds were ob- yielded results nearly identical to those of scored 1–4 (gray to brownish gray) occur served with a score of 9, which is relatively Payne (1974). throughout the South American range of rare in B. virescens (Payne 1974). In this study we chose to score neck col- B. striata, and specimens scored 6–9 (gray- When neck scores were collapsed into oration based on photographs of live birds, ish red-brown to purplish brown) occur three categories (1–3, 4–6, and 7–9), a lower which we considered more objective than throughout the North American range of proportion of intermediate phenotypes (4– scoring neck coloration based on observa- B. virescens. Potential hybrids, especially 6) occurred among our sample of live indi- tions in the field. Our initial assessments of those that have backcrossed with a parental viduals than for specimens collected during neck color scores were nearly always within phenotype, may be difficult to distinguish May–September (33% versus 53%), but the ±1 score of our final consensus. However, 2 from presumably “pure” phenotypes (Hayes differences were not significant (x 2=2.01, variation in lighting and exposure poten- 2006). Individuals scored as 5 occur only in P=0.37; Table 1). The proportions of pheno- tially affects the scoring of neck coloration in the hybrid zones and in isolated B. v. baha- types between our sample of live individuals photographs of live birds. In a few individu- mensis of the Bahamas (Payne 1974); thus, and specimens collected year-round differed als represented in multiple photographs, the 2 individuals with a neck coloration score of significantly (x 2=6.23, P=0.04), but the pro- neck color score varied by ±1 score, in which 5 in Panama presumably represent hybrids, portion of intermediate phenotypes in our case our chosen score was based on the av- whereas those with lower or higher neck col- sample was only slightly lower (33% versus erage of the photographs. We conclude that oration scores may be either relatively “pure” 43%) and contributed only 9% of the overall using Payne’s (1974) voucher specimens to or hybrid genotypes. chi-square value. The significant differences score neck color in Butorides herons is accu- We compared our data on current neck between our sample of live individuals and rate within ±1 score. color variation in central Panama with his- specimens collected year-round is attribut- Our data indicate that all phenotypes torical data published by Payne (1974) for able to proportionately more gray-necked ranging from 1–8 are present during the 44 specimens collected from central Panama phenotypes (scores of 1–3; 52% versus 27%) month of July, at a time when migratory B. (excluding offshore islands) between 79–81° and fewer rufous-necked phenotypes (scores v. virescens from North America is unlikely W longitude, during the period of 1908– of 7–9; 14% versus 30%) in our sample, to be present, and when local populations 1966 (dates from online specimen catalogs). presumably reflecting a lower proportion of Butorides nest with published egg dates Because neck color scores were ordinally (or absence) of migratory B. v. virescens from ranging from 25 February (Wetmore 1965) ranked and did not meet the assumptions North America in our July sample. to 28 July (Van Tyne 1950). As discussed of parametric statistical tests, nonparametric No two individuals were seen within 10 above, the identity of Butorides herons nest-

Volume 67 (2013) • Number 1 5 variation & hybridization in Green Heron and Striated Heron

ing in central Panama has been disputed. Acknowledgments Hellmayr, C. E., and B. Conover. 1948. Van Tyne (1950) carefully identifiedB. stri- Our field work in Panama was funded by Catalogue of birds of the Americas and the ata nesting on Barro Colorado Island and the Margaret Huse Faculty Research Fund of adjacent islands. Field Museum of Natural questioned the credibility of earlier reports the Department of Biology of Pacific Union History, Zoology Series 13, 1(2): 1–434. of nesting B. virescens; thus, it remains un- College. Johnson, N. K., J. V. Remsen, Jr., and C. certain whether B. virescens actually breeds Cicero. 1999. Resolution of the debate in central Panama. Our observations of both Literature cited over species concepts in ornithology: a immature and adult B. virescens during the American Ornithologists’ Union [A.O.U]. new comprehensive biologic species con- breeding season strongly suggest that they 1976. Thirty-third supplement to the cept. Pp. 1470-1482 in: Proceedings of the nest in the area, although some or all may AOU Check-list. Auk 93: 875-879. 22nd International Ornithological Congress, be non-breeding migratory individuals from ----. 1983. American Ornithologists’ Union Durban (N. J. Adams and R. H. Slotow, elsewhere. Check-list of North American Birds. Sixth eds.). BirdLife South Africa, Johannesburg. The increased variability and intermedia- edition. American Ornithologists’ Union, Monroe, B. L., Jr., and M. R. Browning. cy of Butorides herons in central Panama, in Washington, DC. 1992. A re-analysis of Butorides. Bulletin of contrast with populations dominated by B. ----. 1993. Thirty-ninth supplement to the the British Ornithologists’ Club 112: 81-85. virescens in western Panama and populations AOU Check-list. Auk 110: 675-682. Oberholser, H. C. 1916. A revision of the sub- dominated by B. striata in eastern Panama, ----. 1998. American Ornithologists’ Union species of the Green Heron (Butorides vire- strongly implies that hybridization still oc- Check-list of North American Birds. Seventh scens [Linnaeus]). Proceedings of the United curs between the two taxa in central Panama. edition. American Ornithologists’ Union, States National Museum 42: 529-577. Because random mating within a hybrid zone Washington, DC. Ridgely, R. S. 1976. A Guide to the Birds of tends to reduce variability around an inter- Angehr, G. R., and R. Dean. 2010. The Panama. Princeton University Press, mediate phenotype, the current presence of Birds of Panama: A Field Guide. Cornell Princeton, New Jersey. the full range of phenotypes among herons University Press, Ithaca, New York. Payne, R. B. 1974. Species limits and varia- presumably breeding in central Panama sug- Banks, R. C., C. Cicero, J. L. Dunn, A. W. tion of the New World Green Herons gests that most B. virescens and B. striata mate Kratter, P. C. Rasmussen, J. V. Remsen, Butorides virescens and Striated Herons B. assortatively despite frequent hybridization. Jr., J. D. Rising, and D. F. Stotz. 2003. striatus. Bulletin of the British Ornithologists’ We concur with Hayes (2002, 2006) that the Forty-fourth supplement to the American Club 94: 81-88. two taxa appear to have achieved essential Ornithologists’ Union Check-list of North Payne, R. B. 1979. Family Ardeidae. Pp. reproductive isolation (Johnson et al. 1999), American Birds. Auk 120: 923-931. 193–243 in: Check-list of Birds of the thus supporting their current treatment as Corbin, K. W., and G. F. Barrowclough. World. Volume 1. (E. Mayr and G. W. distinct species. 1977. Reproducibility of hybrid index Cottrell, eds.). Harvard University Press, On the southeastern Caribbean island of scores. Condor 79: 497-498. Cambridge, Massachusetts. Tobago, Hayes (2006) demonstrated a his- Eisenmann, E. 1952. Annotated list of birds Schueler, F. W., and J. D. Rising. 1976. torical decline in the proportion of interme- of Barro Colorado Island, Panama Canal Phenetic evidence of natural hybridiza- diate phenotypes (neck color scores of 4–6), Zone. Smithsonian Miscellaneous Collections tion. Systematic Zoology 25: 283-289. which accounted for 72% of 18 specimens 117: 1-62. Stone, W. 1918. Birds of the Panama Canal collected from 1892–1913 but only 34% of Griscom, L. 1929. Birds from Cana, Darien. Zone, with special reference to a collection 50 live individuals during 2000–2002, sug- Bulletin of the Museum of Comparative made by Mr. Lindsey L. Jewel. Proceedings gesting a recent increase in assortative mat- Zoology 69: 147-190. of the Academy of Natural Sciences of ing despite occasional hybridization. Our ----. 1935. The ornithology of the Republic Philadelphia 70: 239-280. data from central Panama suggest a similar of Panama. Bulletin of the Museum of Thayer, J. E., and O. Bangs. 1906. Vertebrata historical decline in the current proportion Comparative Zoology 78: 261-382. from the savanna of Panama. III. Aves. of intermediate phenotypes (33% with neck Hallinan, T. 1924. Notes on some Panama Bulletin of the Museum of Comparative color scores of 4–6) compared with speci- Canal Zone birds with special reference to Zoology 46: 213-224. mens collected year-round 65–103 years their food. Auk 41: 304-326. Van Tyne, J. 1950. Birds notes from Barro earlier (43%) and using a smaller subset of Hayes, F. E. 2002. Geographic variation, hy- Colorado Island, Canal Zone. Occasional specimens collected during May–September bridization, and taxonomy of New World Papers of the Museum of Zoology, University (53%). The lack of statistical significance Butorides herons. North American Birds 56: of Michigan 525: 1-12. could be explained by the more recent 4-10. Wetmore, A. 1965. The Birds of the Republic dates of specimens collected from central ----. 2006. Variation and hybridization in of Panama. Part 1. Tinamidae (tinamous) Panama (compared with Tobago), dilution the Green Heron (Butorides virescens) and to Rynchopidae (skimmers). Smithsonian of the year-round sample by migrant B. vi- Striated Heron (B. striata) in Trinidad and Miscellaneous Collections 150: 1-483. rescens with high neck color scores, and the Tobago, with comments on species limits. Willis, E. O., and E. Eisenmann. 1979. A small number of specimens in the May– Journal of Caribbean Ornithology 19: 12-20. revised list of birds of Barro Colorado September sample. Although it is premature Hayes, F. E., and B. D. Hayes. 2006. First re- Island, Panama. Smithsonian Contributions to conclude that any historical shift in gene cord of Striated Heron (Butorides striata) to Zoology 291: 1-31. frequencies has occurred in central Panama, for the Greater Antilles in St. John, United Zar, J. H. 1998. Biostatistical analysis. Fourth such a pattern should be looked for in fu- States Virgin Islands. North American Birds edition. Prentice-Hall, Inc., Englewood ture decades. 60: 472-473. Cliffs, New Jersey. n

6 north american birds