Australian Field Ornithology 2019, 36, 168–172 http://dx.doi.org/10.20938/afo36168172

Notes on the distribution and plumage variation of Varied and Mangrove

Leo Joseph1*, Jeffrey L. Peters2, Ian J. Mason1 and Alex Drew1

1Australian National Wildlife Collection, National Research Collections , CSIRO, G.P.O. Box 1700, Canberra ACT 2601, Australia 2Department of Biological Sciences, Wright State University, Dayton OH 45435, Ohio, United States of America *Corresponding author. Email: [email protected]

Abstract. Considered conspecific in some recent literature, the Varied versicolor and Mangrove Honeyeaters G. fasciogularis have long been thought to intergrade in and around the city of Townsville, Queensland. Here we update knowledge concerning the distribution of these two species and their variation in plumage colour and pattern in the Townsville region. A lack of records between Lucinda and Townsville seems to reflect a real gap. Further work is required to clarify whether elsewhere Varied Honeyeaters may occasionally exclude Mangrove Honeyeaters in a given locality. If so, this may explain the seemingly inconsistent reports of one or the other at particular localities. Mangrove Honeyeaters immediately south of Townsville may be slightly yellower ventrally than populations further south. However, we argue that this is consistent with our genetic findings reported elsewhere and is not, we conclude, indicative of an undescribed subspecies.

Introduction and ecology, he suggested that they are conspecific but advised further sampling between Cardwell in the north (where to that time only pure Varied Honeyeaters had Varied Gavicalis versicolor and Mangrove Honeyeaters been collected) and Townsville (from where he described G. fasciogularis replace each other latitudinally on intermediates). Noting that populations at Townsville and Australia’s eastern seaboard roughly north (Varied) and Ayr are intermediate in coloration but closest to Mangrove south (Mangrove) of the city of Townsville, Queensland , he reasoned that the zone of intergradation (Ford 1978; Schodde & Mason 1999). The Varied must be centred somewhere north of Townsville and south Honeyeater ranges in eastern Queensland from Cape York of Cardwell. Finally, Ford (1978, p. 73) noted a corollary that Peninsula and Torres Strait south to approximately Lucinda intergradation must be “fairly steep” because the distance and Townsville (Figure 1). It also occurs in New Guinea. between Cardwell and Townsville is only ~150 km. The currently is understood to range from just south of Townsville southwards to Brisbane and Wolstenholme (1925) noted observations of what also northern New South Wales (see Schodde & Mason appeared to be either Varied Honeyeaters or more 1999; Higgins et al. 2001). A narrow hybrid zone is reputedly resembling Varied than Mangrove Honeyeaters at North centred on Townsville in north-eastern Queensland (Ford Keppel Island off Rockhampton ~600 km south-east 1978; Schodde & Mason 1999; Figure 1). Genomic study of Townsville. We shall return to this and other reports of the interactions between the two species in this zone of Varied Honeyeaters from south of Townsville in the of putative hybridisation is the focus of a separate paper Discussion. (Joseph et al. 2019). Schodde & Mason (1999, p. 237) argued that a slightly Ingram (1908) reported that Mangrove Honeyeaters from larger series of specimens than had been available to Inkerman (~100 km south of Townsville) have a paler ventral Ford (1978) affirmed that evidence for what they termed surface relative to populations further south. Later, Ford “character flow” (presumably meaning gene flow) is patchy (1978) examined five specimens that had been recently and localised. They retained the Varied and Mangrove collected between Townsville and ~100 km further north at Honeyeaters as separate species. They also noted the Ingham and other comparative material (unspecified) from patchy occurrence of these species between Halifax Bay the American Museum of Natural History, New York. Most of in the south and Townsville in the north. Molecular studies the latter was important comparative material of Mangrove (Nyari & Joseph 2011; Joseph et al. 2014; Marki et al. and Varied Honeyeaters from beyond the Townsville 2017) have since shown that these two species are not region of interest here. Ford (1978) argued that specimens each other’s closest relatives; contrary to expectations from Cleveland Bay, immediately south of Townsville, and based simply on their putative hybridisation, the Mangrove Inkerman, further south, were phenotypically intermediate Honeyeater and G. virescens, which between Varied and Mangrove Honeyeaters, particularly is widespread in inland Australia, are each other’s closest on the ventral surface. He included a photograph of one relatives, and the is the closest relative of those specimens aligned with specimens more typical to that pair of species. of Varied and Mangrove Honeyeaters. Ford’s (1978) The purpose of this note is to complement a separate conclusions were cautiously worded. For example, genomic study of contact between the two species (Joseph he noted that Varied and Mangrove Honeyeaters are et al. 2019). We briefly update knowledge on distribution connected by a population of intermediate coloration in and geographical variation in plumage colour and the region where it has been considered that they come patterning in these birds. Our scope for the latter focused in contact. When coupled with similarities in vocalisations on Mangrove Honeyeaters as a result of our genomic Varied and Mangrove Honeyeaters: Distribution and plumage variation 169

Figure 1. Left: Map showing approximate locations (orange-filled circles) and corresponding GPS coordinates (WGS-84) of locations of unsuccessful searches for Mangrove and Varied Honeyeaters by AD and IJM in May 2015 between Lucinda and Townsville (black squares). Right: Two specimens from the Australian National Wildlife Collection (ANWC) collected in May 2015 from the northern and southern side of Townsville, respectively, a Varied Honeyeater ANWC B57007 (left) and a Mangrove Honeyeater ANWC B57006 (right) separated by a straight-line distance of <5 km.

findings. We hope that the study reported here will guide Plumage any further quantitative study of plumage variation in this species. We qualitatively assessed plumage colour and patterning through admittedly preliminary examination of all specimens held in the Australian National Wildlife Collection (ANWC). Methods We assessed variation in colour particularly on the underparts where we focused on colour and pattern of the throat, chest and belly. Distribution

In May 2015, AD and IJM searched 18 localities Results specifically for Varied and Mangrove Honeyeaters north and south of Townsville, from Lucinda (Taylors Beach and Toomulla Road south of Lucinda) to just south of Distribution Townsville (Figure 1). Recordings of vocalisations (Varied Online resources indicate records of Mangrove or Varied or Mangrove Honeyeater, depending on whether north or Honeyeaters from several localities in and around Lucinda south of Townsville, respectively) were played to attract at the north of the zone of interest. Apart from the many birds. Responses, when they occurred, were almost records in Lucinda itself, almost all such records were from always immediate. In the absence of a response, the historical literature and of imprecise and indeterminate observers departed from the locality within 5 minutes. locality (e.g. “Queensland”). One, a specimen record We also examined literature sources (Ford 1978; Ford from Ingham (Queen Victoria Museum and Art Gallery, et al. 1981; Storr 1984; Wieneke 2000; Nielsen 2015) and Launceston, Tasmania, 2006:2:15550), is a single- online resources (Atlas of Living Australia https://www. egg clutch, collected from a nest in a drooping branch ala.org.au/ and eBird https://ebird.org/home; accessed ~1 m above the water level on 20 September 1961 by 12 July 2019) for distributional records. Between 1978 and H. Gatebridge (D. Maynard pers. comm.). Nielsen (2015, 1999, Mangrove and Varied Honeyeaters were considered p. 339) referred to a “large population” of Mangrove conspecific so observers before the 2000s tended not to Honeyeaters about Lucinda, which is based on sightings differentiate between the two, simply recording them as by John Young between c. 1979 and 1996 (L. Nielsen pers. Varied Honeyeaters. This means that caution is warranted comm. 4 April 2018). Conversely, Nielsen (2015, p. 339) in interpreting literature records. also noted the Ross River further south in Townsville as 170 Australian Field Ornithology L. Joseph et al.

Figure 2. Representative series of Mangrove Honeyeaters from Townsville, Queensland (left) to northern New South Wales (right). All are held in the ANWC. From left to right, with abbreviated localities and the B prefix of the registration numbers omitted for brevity: 57006 (Townsville), 29009-29017-29010 (1 km S of Ross River mouth), 56984-31208-28779-28780 (25–30 km SE of Ayr), 31192-28774-31191-31193 (40 km SE of Ayr), 57004-57005 (5 km SE of Bowen), 39469 (SE of Proserpine), 39473-39470 (Cape Conway Peninsula), 31130-28770-31129-28769 (N of Mackay), 39253 (Tweed Heads), and 39339 (N of Bundaberg). the “dividing line” between the two species (D. James in were 4.6 km apart in a straight line and separated essentially Nielsen 2015). by urban Townsville itself (Figure 1); LJ revisited the location of the Mangrove Honeyeater specimen in August Notably at one locality, Forrest Beach (–18.71, 146.27) 2015 and recorded only Varied Honeyeaters, identifying south of Lucinda, at least one observer confidently them by their entirely yellow underparts and relatively large recorded Mangrove Honeyeaters, at an unspecified time size. We note that the northernmost ANWC specimens of before 1978, well north of their current range and well into Mangrove Honeyeaters are from 1 km south of the Ross that of the current range of Varied Honeyeaters (D. Seton River mouth, Townsville (ANWC B29009–29011, June in Ford 1978). However, other observers have specifically 2000). searched for either species at that same locality and noted only their absence (e.g. Ford 1978; present study; AD & IJM pers. obs.). There are no records of either Plumage Variation Mangrove or Varied Honeyeaters between Forrest Beach (just south of Lucinda) and Bushland Beach (–19.20, 146.68; just north of Townsville). Further south, closer to Examination of the 22 northern and 17 southern specimens but still north of Townsville, at Rollingstone, both species of Mangrove Honeyeaters available to us at the ANWC or have been recorded (Officer 1964); Ford (1978) queried subsets of them (Figures 1 and 2) suggests three minor and subtle differences in plumage between northernmost whether these might have been referable to intermediate populations (just south of Townsville to Bowen) and those phenotypes, which he felt would be difficult to distinguish further south. Relative to the southernmost populations of from either species. Again, in the present study, we found Mangrove Honeyeaters (Figure 2), the northernmost birds none of these birds near Rollingstone (AD & IJM pers. tend to show brighter yellow throat barring, lighter upper obs. May 2015). We cannot assess the reliability of sight ventral (chest) coloration because of the more prominent records of Varied Honeyeaters south of Townsville and grey streaking and thus absence of the brownish chest- Mangrove Honeyeaters at or north of Lucinda and note band sometimes apparent in southern birds, and more again that many observers will simply have recorded extensive grey on the lower ventral surface (belly) perhaps Varied Honeyeaters when the two forms were treated as because of less streaking. Individual variation is marked, conspecific. however. Similarly, we recorded no birds between Taylors Beach south of Lucinda and Rowes Bay in the northern suburbs of Townsville itself (Figure 1). Most notably, AD and IJM Discussion (present study, May 2015) collected one Varied Honeyeater (ANWC B57007) and one Mangrove Honeyeater (ANWC We set out to provide a brief update concerning B57006) at Rowes Bay and South Port Road, southern distribution and plumage variation in Mangrove and Varied suburban Townsville, respectively. These sampling sites Honeyeaters especially where they approach each other Varied and Mangrove Honeyeaters: Distribution and plumage variation 171

and interact genetically (Joseph et al. 2019) around the way gene flow from Varied Honeyeaters into Mangrove city of Townsville. A surprising gap of records of any of Honeyeaters that we suggest explains the subtle but less these birds between Lucinda and Townsville is striking, yet than consistent phenotypic and genotypic distinctiveness suitable habitat seemed to be present in our surveys. of some northernmost Mangrove Honeyeaters. A challenge for later phenotypic analyses is that of undertaking accurate Distributional overlap and replacement of one species by the other on fine geographical and temporal scales, quantitative measurements of key characters such as such as in and around Townsville itself, is supported by intensity of ventral streaking and the ground colour of the three lines of evidence. First are our own observations chest plumage using methods such as spectrophotometry. of one or the other, but not both, species at particular Finally, we return to earlier reports of Varied Honeyeaters localities in and around Townsville in 2015. Second are south of Townsville to as far as the Keppel Island Group reports of the Ross River being the limit between the (e.g. Wolstenholme 1925; Nix 1972; Catling et al. 1992). two species (D. James in Nielsen 2015) whereas others Higgins et al. (2001) considered all of these doubtful. have in earlier decades recorded Mangrove Honeyeaters For example, Nix (1972) reported both Varied and well to the north. Third, there is variation over time as Mangrove Honeyeaters in the Shoalwater Bay area but to whether any of these birds are even present at some referred only to having heard vocalisations when listing localities especially between Lucinda and Townsville, Varied Honeyeaters. Unfortunately, this is insufficient two examples being near Townsville at Rollingstone and to allow unequivocal acceptance of both species being Forrest Beach. Possibly relevant to explaining this is that present. Several further points might usefully be raised Varied Honeyeaters are, of course, substantially larger in this regard. One is that we should indeed be alert to than Mangrove Honeyeaters (Schodde & Mason 1999). the possibility that introgression of the yellow ventral Given that food sources for these birds are also spatio- tones of Varied Honeyeaters into Mangrove Honeyeaters temporally variable, we suggest that Varied Honeyeaters may occasionally be observed much further south than may competitively exclude Mangrove Honeyeaters from a the very limited extent suggested in our genomic study given locality. This could result in either of the two species, (Joseph et al. 2019). A second, more important point is but not necessarily both, being present on different days in the legacy of the 22 years between 1978 and 1999 when areas such as Townsville. Clearly, and perhaps despite this, Varied and Mangrove Honeyeaters were considered to the genomic data (Joseph et al. 2019) suggest that the two be two subspecies of one species, the Varied Honeyeater species have occasionally interbred. Carefully conducted G. versicolor. In that period, if one were to report the monitoring at specific localities between Townsville and Mangrove Honeyeater at the species level one would Lucinda would likely be rewarding here. have listed it as a Varied Honeyeater. In the absence of Although individual variation in plumage colour and specimens, photographs or detailed accompanying notes patterning among individual Mangrove Honeyeaters is describing the birds involved, this opens a source of marked, a quantitative assessment is necessary to even confusion as to what had actually been seen. Specimens confirm that there is geographic and not simply individual held in the ANWC and that were collected during the course variation. Pending such a study, we here provide the of work reported by Catling et al. (1993) make it clear that following preliminary assessment. those authors assigned the name Varied Honeyeater to Newly available specimens that we have examined their specimens of the Mangrove Honeyeater. Coupled with raise the possibility of hitherto unrecognised differentiation the fact that Varied Honeyeaters also occur in mangroves, between northern and southern populations of the it may also explain likely confusion underpinning some Mangrove Honeyeater. What is little more than a hint of reports of both species from the 1980s and even in the last geographically structured variation is apparent when few years in eBird.org (accessed 14 October 2019). These viewing a series of specimens, not one or two individuals, concern Varied Honeyeaters south of Townsville (from the because of confounding effects of individual variation last few years) and, arguably, Mangrove Honeyeaters north (Figure 2). A similar situation is apparent in the Mulga of Cairns (from 1984). Third, the yellower appearance of Parrot Psephotellus varius (see colour plate in McElroy et some Mangrove Honeyeaters that we have reported here al. 2018). It parallels the evidence of similarly slight but from near Townsville and explained through our genomic latitudinally structured genomic variation within Mangrove work (Joseph et al. 2019) may also lead some observers Honeyeaters that we have reported elsewhere (Joseph to report Varied Honeyeaters. et al. 2019). The division between northern and southern A final point concerns Wolstenholme’s (1925) more populations for genomic characters, as for plumage, is compelling report of true Varied Honeyeaters or, at least, suggested in the Bowen area at ~21°S in latitude. We reject far more Varied-like birds at North Keppel Island, ~600 km the possibility that there is an undescribed subspecies of south-east of Townsville. Although especially interesting, Mangrove Honeyeater in the northernmost part of its range. his report is clearly anomalous in terms of more recent We favour the alternative explanation that any yellower sightings, where, for example, only Mangrove Honeyeaters appearance of the northern populations is consistent with have specifically been recorded there (e.g. eBird.org our genomic findings, i.e. low-level, one-way gene flow accessed 14 October 2019). from Varied into Mangrove Honeyeaters near Townsville (Joseph et al. 2019). We consider that Ford (1978) correctly We conclude that detailed study of the scale of movements interpreted five Cleveland Bay specimens from just south of these two species, especially the Varied Honeyeater, of Townsville as intergrades, i.e. intermediate in plumage and their ecological interactions would be rewarding. This between Varied and Mangrove Honeyeaters. Equally, he would usefully include modelling of their past distributions. reasonably erred in inferring a hybrid zone having two-way Changing sea levels since the Last Glacial Maximum gene flow between the two species. Similarly, Schodde c. 20,000 years ago may bring new perspectives on how & Mason’s (1999) interpretation is consistent with one- their present-day distributions have themselves evolved. 172 Australian Field Ornithology L. Joseph et al.

Acknowledgements Marki, P.Z., Jønsson, K.A., Irestedt, M., Nguyen, J., Rahbek, C. & David Maynard and Judy Rainbird, Queen Victoria Museum and Fjeldså, J. (2017). Supermatrix phylogeny and biogeography of the Australasian Meliphagides radiation (Aves: Passeriformes). Art Gallery, Launceston, Tasmania, provided data from specimens Molecular Phylogenetics and Evolution , 516–529. held in their care. CSIRO Global and National Collections and 107 McElroy, K., Beattie, K., Symonds, M. & Joseph, L. (2018). Marine Infrastructure provided funding. We thank Richard Noske, Mitogenomic and nuclear diversity in the Mulga Parrot of the James Fitzsimons and Hugh Ford for their helpful critiques of Australian arid zone: Cryptic subspecies and tests for selection. the submitted manuscript and for alerting us to some important, Emu 118, 22–35. additional literature. Nielsen, L. (2015). Birds of the Wet Tropics of Queensland & Great Barrier Reef & Where To Find Them. Author, Mount Molloy, Qld. References Nix, H.A. (1972). Fauna of the Shoalwater Bay area. In: Gunn, R.H., Galloway, R.W., Walker, J., Nix, H.A., McAlpine, J.R. Catling, P.C., Mason, I.J., Richards, G.C., Schodde, R. & & Richardson, D.P. (Eds). Shoalwater Bay Area, Queensland. Wombey, J.C. (1992). The Land Vertebrate Fauna of the Technical Memorandum 72/10, pp. 107–125. Division of Land Eastern Dunefields and Tidal Zone, Shoalwater Bay Training Research, CSIRO, Canberra. Area, Queensland. Final Report – November 1993. CSIRO Nyari, A. & Joseph, L. (2011). Systematic dismantlement Division of Wildlife & Ecology, Canberra. of improves the basis for understanding Ford, J. (1978). Intergradation between the Varied and Mangrove relationships within the honeyeaters (Meliphagidae) and Honeyeaters. Emu 78, 71–74. historical development of Australo-Papuan communities. Ford, J., Greensmith, A. & Reid, N. (1981). Notes on the Emu 111, 202–211. distribution of Queensland birds. Sunbird 11, 58–69. Officer, H.R. (1964). Australian Honeyeaters. Bird Observers Higgins, P.J., Peter, J.M. & Steele, W.K. (Eds) (2001). Handbook Club, Melbourne. of Australian, New Zealand & Antarctic Birds, Volume 5: Tyrant- Schodde, R. & Mason, I.J. (1999). Directory of Australian Birds, flycatchers to Chats. Oxford University Press, Melbourne. Volume 1: . CSIRO Publishing, Melbourne. Ingram, C. (1908). On the birds of Inkerman Station, North Storr, G.M. (1984). Revised list of Queensland birds. Records of Queensland. Ibis 50, 458–481. the Western Australian Museum Supplement 19, 1–92. Joseph, L., Drew, A., Mason, I.J. & Peters, J.L. (2019). Wieneke, J. (2000). Where to Find Birds in North-east Queensland. Introgression between non-sister species of honeyeaters Author, Townsville, Qld. (Aves: Meliphagidae) several million years after speciation. Wolstenholme, H. (1925). Notes on the birds observed during the Biological Journal of the Linnean Society 128, 583–591. Queensland congress and camp-out, 1924. Emu 24, 243–251. Joseph, L., Toon, A., Nyári, Á.S., Trueman, J. & Gardner, J. (2014). A new synthesis of the molecular systematics and biogeography of honeyeaters (Passeriformes: Meliphagidae) highlights biogeographical complexity of a spectacular avian Received 29 July 2019, accepted 14 October 2019, radiation. Zoologica Scripta 43, 235–248. published online 6 December 2019