The analysis of growth in extinct organisms is difficult. The general lack of skeletal material from a range of developmental states precludes determination of growth characteristics. For New Zealand's extinct moa we have available to us a selection of rare femora at different developmental stages that have allowed a preliminary determination of the early growth of this giant flightless bird. We use a combination of femora morphometrics, ancient DNA, and isotope analysis to provide information on the identification, classification, and growth of extinct moa from the genus Euryapteryx.
Using ancient DNA, we identify a number of moa chick bones for the species Euryapteryx curtus, Dinornis novaezealandiae, and Anomalopteryx didiformis, and the first chick bone for Pachyornis geranoides. Isotope analysis shows that ∂15N levels vary between the two known size classes of Euryapteryx, with the larger size class having reduced levels of ∂15N. A growth series for femora of the two size classes of Euryapteryx shows that early femora growth characteristics for both classes are almost identical. Morphometric, isotopic, and radiographic analysis of the smallest Euryapteryx bones suggests that one of these femora is from a freshly hatched moa at a very early stage of development.
Using morphometric, isotopic, and ancient DNA analyses have allowed the determination of a number of characteristics of rare moa chick femora. For Euryapteryx the analyses suggest that the smaller sized class II Euryapteryx is identical in size and growth to the extant Darwin's rhea.
Citation: Huynen L, Gill BJ, Doyle A, Millar CD, Lambert DM (2014) Identification, Classification, and Growth of Moa Chicks (Aves: Dinornithiformes) from the Genus Euryapteryx. PLoS ONE 9(6): e99929. doi:10.1371/journal.pone.0099929
Editor: Tom Gilbert, Natural History Museum of Denmark, Denmark
Received: February 18, 2014; Accepted: May 20, 2014; Published: June 12, 2014
Copyright: © 2014 Huynen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors thank the Australian Research Council for support (Grant number DP110101364; The molecular evolution of wings in ratites). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
In depth analysis of growth in ancient animals is often limited due to the scarcity and degraded nature of skeletal material or tissues of different ages. Similarly, the rare occurrence of different aged bones for New Zealand's extinct ratite moa (Aves: Dinornithiformes) has made any analysis of moa growth difficult , , .
Adult moa ranged in size from less than 20 kg for the small coastal moa Euryapteryx curtus curtus to over 200 kg for the South Island giant moa Dinornis robustus . The identification of species within the Euryapteryx genus has been particularly difficult. Latest data suggest the existence of a small subspecies (E. curtus curtus) limited to New Zealand's North Island, and a larger subspecies (E. curtus gravis) found only in New Zealand's South Island .
How moa grew is largely unknown with most published work comparing moa to the growth characteristics of their extant relatives , . Relatively recent work analysing cortical growth marks in moa limb bones suggest that, unlike their modern relatives, moa had a particularly long pre-adult growth period .
We analyse moa growth using a number of rare moa chick femora, currently housed at New Zealand's Auckland Museum and kindly made available to us. The museum houses a significant number of samples of moa chick bones from sand-dune sites in New Zealand's upper North Island, especially from the Karikari Peninsula/Doubtless Bay area, including Tokerau Beach. Adult bones from these sites have been attributed to three moa species with most being derived from E. curtus curtus , , .
To date, only one embryonic moa has been identified to species, where bones associated with an egg were shown to belong to the heavy-footed moa Pachyornis elephantopus , . As bones of developing chicks, which often lack identifying characters, are particularly difficult to identify ,  we use a minimally destructive technique to genetically assign differently sized immature moa femora to the species level. We then use bone morphometrics to present a growth series of chick femora for Euryapteryx. In addition, we present isotope and radiographic data for the smallest moa femora to determine whether these may have derived from unhatched eggs. The isotope data has also allowed us to further explore the status of two subspecies proposed for moa from the genus Euryapteryx .
We successfully amplified a relatively short (~70 bp) hypervariable mitochondrial DNA fragment from 29 of 32 immature bones sampled from various locations in New Zealand (Tables 1,2; Figures 1,2). Femur LB6261c was identified as belonging to Dinornis novaezealandiae and is 72 mm long (Table 2, Figure 2). Bones of a late-term embryonic moa (identified as Pachyornis elephantopus) were recovered from inside an egg in 1866 , . The egg was 226 mm long and 155 mm wide with the embryonic femur being approximately 48 mm long (with ends restored). An egg found at Kaikoura and attributed to Dinornis was shown to be 240 mm long, and by proportion, its embryo (if at the same stage as the Pachyornis egg) would have had a femur approximately 51 mm long. Therefore the size of LB6261c suggests it was from a recent hatchling. Further evidence for the extreme immaturity of this femur is the lack of caudal tuberosities on the femur shaft, a feature of the femur that separates Dinornis from the emeid moas , . Turvey and Holdaway (2005)  described ‘postnatal’ bones of Dinornis and showed that femora at growth-stage 1 began at 156 mm in length with their youngest stage 1 individual having an estimated weight of 15.8 kg. Thus their sample included only well-grown chicks and did not include hatchlings. The single Pachyornis geranoides femur: At 82 mm long, LB7976 (Table 2) is likely to be from a well-developed chick, since this moa species is relatively small. Although difficult to determine due to erosion, an excavation at the proximal end of the bone may be the pneumatic fossa that characterises this species , . LB7976 is the only known chick bone of this species. Four Anomalopteryx didiformis femora were identified; AIM LB6666c, AIM LB6261a-b, and AIM LB6285a (Table 2), all derive from Doubtless Bay, and are the first chick bones to be identified (by DNA) for this species, as well as being the first record of this species from Doubtless Bay. For Euryapteryx curtus 22 femora specimens were identified by DNA, and include the 11 smallest femora (see Figure 2). These form the first large sample of chick bones attributable with certainty to this species, and were used for detailed morphological, radiographic, and isotopic analyses. The sequence targeted allows discrimination of all moa species and identifies two distinct genetic haplotypes for Euryapteryx. The haplotypes can be separated by a single SNP that associates class I Euryapteryx with thick eggshells (0.98 mm - 1.60 mm) and class II Euryapteryx with thin eggshells (0.74 mm–0.98 mm) . The distribution of the two Euryapteryx classes (I and II) closely mimics the distribution proposed for subspecies E. curtus gravis and E. curtus curtus respectively .
Figure 1. Moa sample locations.
The locations of moa samples used in this work are shown.doi:10.1371/journal.pone.0099929.g001
Figure 2. Examples of moa chick femora.
From left to right: LB5990, LB8295, LB6070, LB6261d, LB6284, LB6261c, LB12961, LB6285b, LB6657, LB6071, LB6069. All are left femora except the two at far right. DNA analysis suggests all are from Euryapteryx except for LB6261c which was identified as Dinornis. Gridlines are at 10 mm intervals.doi:10.1371/journal.pone.0099929.g002
Table 1. Moa sequences.doi:10.1371/journal.pone.0099929.t001
Table 2. Identification and morphometrics of moa chick femora from Auckland Museum.doi:10.1371/journal.pone.0099929.t002
The length and width of each femur was recorded for both Euryapteryx classes and compared to obtain a growth series (Figure 3). No difference could be found in early femora growth characteristics between the two Euryapteryx classes. Morphometric and mass calculations were carried out on the smallest Euryapteryx femur (LB12960; belonging to class II Euryapteryx) to determine whether this femur may have been from an unhatched egg. Two femur-based equations exist that are commonly used to determine avian mass; one using least shaft circumference provided by  and one using total femur length  (see methods). Although both methods are relatively accurate for the determination of mass of an avian adult, mass calculation using femur length has been shown to be more accurate for developing birds . Using femur length, estimated at c. 47 mm (measured at 44 mm but adding 3 mm to allow for missing ends) the mass of the LB12960 individual was calculated to be only 470 g. For Darwin's rhea, very similar to class II Euryapteryx in adult size (15–28.6 kg) and eggshell thickness (0.73–1.1 mm)  newly hatched individuals range in weight from 0.327–0.491 kg with an average weight of 0.426 kg . This suggests that femur LB12960 may have derived from a moa embryo, but is likely to be from a very young newly hatched chick.
Figure 3. Femoral growth series for chick bones from Euryapteryx curtus.
Graph showing femur length vs width (mm). Blue - Euryapteryx class I, eI; Red - Euryapteryx class II, eII; For comparison, chick bones from other moa species identified are also shown. Green - D. novaezealandiae, Dn; Purple - A. didiformis, Ad; Brown - P. geranoides, Pg. A line of best fit is shown in grey.doi:10.1371/journal.pone.0099929.g003
Further analysis of the developmental stage of the smallest Eurypateryx femora was carried out by radiography. Six bones were analysed, of which four had essentially intact bone of variable thickness at the ends (Table 3). These ends correspond to the metaphyses of the femora adjacent to the growth plates, the epiphyses having been lost or separated. The ends of intact bones suggest that the cartilage ‘cones’ described by  had already migrated away from the bone ends and growth plates. This is a feature found in their study of rheas only after the birds were of 3 weeks maturity or more. It is therefore reasonable to assume that these four chicks were a few weeks old at least. Only two of the moa chick bones had complete defects in the metaphyses, identifying them as neonatal or embryonic (Figure 4). One of these (LB12960) has a slightly larger defect than the other and subjectively has less trabecular bone overall than any of the others. These data further suggest that this bone may have come from a late stage embryo or newly hatched chick, in concordance with the nitrogen isotope (below) and morphometric findings (Table 2).
Figure 4. Radiographic analysis of Euryapteryx class II femora.
A. Chick LB13978: more mature form with intact bone end on right, consistent with chick of a few weeks age where cartilage cone has migrated down shaft. B. Chick LB12960: immature bone with open end on right, consistent with cartilage cone present up to growth plate as occurs with embryo and neonate.doi:10.1371/journal.pone.0099929.g004
Table 3. Radiographic analysis of moa chick femora.doi:10.1371/journal.pone.0099929.t003
Additional analyses were carried out on the smallest femora using isotope counts in the hope that femora from unhatched individuals, by feeding on yolk, have different ∂15N and/or ∂13C levels to femora from hatched chicks feeding on insects. To account for habitat or species-specific biases, isotope levels were determined for several moa species from different environments including sand-dune, cave, and swamp (Table 4, Figure 5). No species-specific differences in isotope levels were found among the few samples that were measured. However, moa bone samples from caves had very low ∂15N values (−0.1–2.90 parts per million (o/oo)), while those from swamps had ∂15N levels of 4.2–8.6 o/oo and sand-dune samples ranged in ∂15N from 3.1–5.5 o/oo, values that are likely to be indicative of the specific environment (Table 4, Figure 5). Isotopic analysis showed relatively low values of ∂13C (18.3–23.1 o/o) in all Euryapteryx bones suggesting that individuals of this moa species preferred more open habitats and, as expected for New Zealand, a C3-based diet .
Figure 5. Isotope analysis of selected moa bones.
Moa bone samples from different species (Table 3) and environments were subjected to isotopic analyses. Light blue - cave samples; Orange - Dune samples; Green - Swamp samples; Blue - Euryapteryx class I, eI; Red - Euryapteryx class II, eII; The isolated eII sample is the possible newly hatched Euryapteryx class II individual.doi:10.1371/journal.pone.0099929.g005
Table 4. Isotope analysis of moa bones from different environments.doi:10.1371/journal.pone.0099929.t004
As almost all Euryapteryx chick bones in this study were obtained from the same area, climate and habitat differences are unlikely to be a source of isotope variation. Interestingly, ∂15N levels between the two Euryapteryx classes showed class I ∂15N values were generally low and ranged from 2.2–3.6 o/oo (n = 3, mean = 2.83, SD = 0.58) while class II ∂15N values ranged from 3.5–5.5 o/oo (n = 6, mean = 4.86, SD = 0.71). Although the sample numbers are low, the class I and class II∂15N values proved to significantly different; p = 0.0115, Table 4, Figure 5). An unusually high ∂15N level (10.3 o/oo) was found for the smallest Euryapteryx bone, AIM LB12960 (Table 4, Figure 5).
Morphometrics of femora from each variant showed very similar growth curves at the early stages of development. For the closely related emu and ostrich it has been shown that cortical bone thickness remains constant for the first two months after hatching, with the highest radial growth rate occurring 7–14 days posthatch . Similarly for Euryapteryx, extrapolation of the growth curve to 0 suggests that at the very early stages of development, femur width increased proportionately more than femur length (Figure 3). For class II Euryapteryx, femur growth is linear until femur length reaches approximately 150 mm.
Isotope analysis gives an indication of diet, climate, and habitat. Carbon 13 (13C) levels in bone give an indication of whether the plants ingested utilized a C3 or C4 photosynthetic pathway, while Nitrogen 15 (15N) levels provide information on the organisms trophic level . In general, bone ∂13C values become more negative when animals feed in open shrubland, as opposed to vegetation from closed canopy or forested areas , , . ∂15N levels are an indication of not only trophic level but can also be indicative of habitat, climate, and salinity, with high ∂15N values being more characteristic of individuals found in more saline, hot, dry environments . The Isotope data we obtained provides further evidence that the two genetic variants of Euryapteryx curtus are likely to represent subspecies. One of these was likely to be a small sized moa that was associated with thin eggshells. The other was probably a larger sized moa associated with thick eggshells . The smallest Euryapteryx class II bone sample (LB12960) returned a very high ∂15N value. High ∂15N values are generally associated with high trophic levels. It has been found, for sharks at least, that embryos have elevated ∂15N levels compared to their mothers . Furthermore ∂15N values similar to those found in sample LB12960 have been obtained for egg yolk from insect-fed (high protein) chickens  suggesting, in accordance with the morphometric and radiographic data, that this femur may have come from inside an egg or from a newly hatched individual.
Newly hatched ostrich and emu chicks increase in body mass by at least 50% per month, and reach adult size within one year . For Euryapteryx, this process is likely to be much slower with the smaller Euryapteryx curtus curtus (North Island subspecies; adult weight ~20 kg;  reaching maturity after 4 years and the larger Euryapteryx curtus gravis (mainly South Island subspecies; adult weight ~80 kg; ) not reaching adulthood for over 9 years . Assuming E. c. curtus and E. c. gravis neonates average 0.43 kg  and 0.95 kg respectively (latter weight based on average ostrich hatch weight; ), this would suggest a growth rate for Eu. curtus curtus of ~10% per month and Eu. curtus gravis a very slow ~5% per month. For this reason, moa chicks were likely to be highly dependent on adult(s) support for a significant period of time. This is somewhat unusual for ratites where extant ratites such as the kiwi, emu, ostrich, and rhea are considered precocious and require little adult support beyond the initial growth stages . In direct contrast to emu, ostrich, and rhea which are subject to predation by goannas, hyenas, and large cats respectively  it is perhaps fortunate that prior to the arrival of humans, moa had very few natural predators . This scarcity of natural predators may have allowed moa to survive its protracted early stages of development.
We use a variety of analytical methods to gain information on rare chick bones from New Zealand's moa. The analyses have allowed the identification of the first chick bone for P. geranoides, the determination of early growth for moa from the genus Euryapteryx, the characterization of a very rare newly hatched Euryapteryx individual, and have also provided futher evidence for the existence of possible subspecies within this genus.
Materials and Methods
Moa bone samples
All moa bone samples were kindly loaned by the Auckland War Memorial Museum (AIM) and Canterbury Museum (CM). Chick bones were identified from the Auckland War Memorial Museum (AIM) moa bone collection by their reduced size in relation to adult bones. Well preserved specimens with known collection locality were selected for sampling. Where possible the left femur was chosen. Permission to sample moa specimens was obtained from the respective museum curators. No permits were required for the described study, which complied with all relevant regulations.
Maximum femur length was determined by measuring parallel to the long axis of shaft, as shown in Figure 4Al of . Femur width was measured at mid-point along the shaft, at right angles to the anterior-posterior plane with the femur head to one side as shown in Figure 5.III.D of . Immature femora were measured irrespective of incompleteness due to a lack of fused epiphyses. Measurements were made to the nearest 0.1 mm using Vernier calipers. Moa weight was determined using total femur length where body mass (kg) = (femur length (mm)/61.64)2.7855 according to .
The smallest moa bones were subjected to computed tomography at high resolution using a multi-detector scanner (Siemens Emotion 16, Siemens Medical, Erlangen, Germany). Helical thin slices (0.6 mm, 90 mm field of view, 130 kilovolt peak, 60 milliampere seconds, 512×512 matrix, bone algorithm) were generated and reconstructed into 0.6 mm contiguous slices along and at right angles to the bone. These were viewed and measured on a PC using Siemens software (Syngo fastView, Siemens Berlin, 2009).
Isotope analysis was carried out on selected bones at the Australian Rivers Institute, Griffith University, Australia using the following standards: Primary; N - Ambient air IAEA-305a, C - ANU sucrose, Elemental; Acetanilide Working; ‘Prawn’. Mass Spectrophotometry was carried out on a GV Isoprime spectrophotometer (Manchester UK) using a Eurovector EA 3000 inlet. Samples were taken from the outer (radial) layer where possible .
DNA extraction and amplification
DNA was extracted, amplified, and sequenced using the mitochondrial primers mcrshFF and mcrshRR as outlined in . DNA was extracted from approximately 50 mg of bone shavings by incubation overnight at 56°C with proteinase K, and then purified by phenol:chloroform extraction and silica bed binding using a Qiagen DNeasy Blood & Tissue Kit. DNA was eluted from the silica column and subjected to PCR using the mitochondrial control region primers mcrshFF and mcrshRR as outlined in  and visualised by agarose gel electrophoresis and ethidium bromide staining. Positive amplifications were sequenced using primer F6t as described in . DNA results were obtained blind, where bone identity remained unknown until sequences were obtained.
Ancient DNA procedures
DNA was extracted and amplified according to the criteria proposed by . Ancient DNA was extracted in a dedicated ancient DNA laboratory at Griffith University, Australia, and amplified at a separate isolated modern lab facility. For sequence verification, several samples were replicated independently at the Massey University Ancient DNA facility, Albany, Auckland, New Zealand.
Conceived and designed the experiments: LH BJG CDM DML. Performed the experiments: LH BJG AD. Analyzed the data: LH BJG AD. Contributed reagents/materials/analysis tools: AD BJG DML. Wrote the paper: LH BJG AD DML.
- 1. Worthy TH (1988) An illustrated key to the main leg bones of moas. National Museum of NZ misc series 17: 3–36.
- 2. Worthy TH, Holdaway RN (2002) The Lost World of the Moa. Prehistoric Life of New Zealand. Bloomington, Indiana, Indiana University Press.
- 3. Turvey ST, Holdaway RN (2005) Postnatal ontogeny, population structure, and extinction of the giant moa Dinornis. Journal of Morphology 265: 70–86.
- 4. Worthy TH, Scofield RP (2012) Twenty-first century advances in knowledge of the biology of moa (Aves: Dinornithiformes): a new morphological analysis and moa diagnoses revised. New Zealand Journal of Zoology 39: 87–153.
- 5. Turvey ST, Green OR, Holdaway RN (2005) Cortical growth marks reveal extended juvenile development in New Zealand moa. Nature 435: 940–943.
- 6. Gill BJ (2010) Regional comparisons of the thickness of moa eggshell fragments (Aves: Dinornithiformes). Records of the Australian Museum 62: 115–122.
- 7. Huynen L, Gill B, Millar CD, Lambert DM (2010) Ancient DNA reveals extreme egg morphology and nesting behaviour in New Zealand's extinct moa. Proc Natl Acad Sci USA 107: 16201–16206.
- 8. Hector J (1872) On recent moa remains in New Zealand. Transactions and Proceedings of the New Zealand Institute 4: 110–120.
- 9. Worthy TH (1987) Sexual dimorphism and temporal variation in the North Island moa species Euryapteryx curtus (Owen) and Pachyornis mappini Archey. National Museum of New Zealand Records 3: 59–70.
- 10. Steadman DW (1980) A review of the osteology and paleontology of turkeys (Aves: Meleagridinae). Contributions in Science of the Natural History Museum of Los Angeles County 330: 131–207.
- 11. Prange HD, Anderson JF, Rahn H (1979) Scaling of skeletal mass to body mass in birds and mammals. The American Naturalist 113: 103–122.
- 12. Holdaway RN, Hawke DJ, Bunce M, Allentoft ME (2011) Identification of an optimal position for stable isotopic analysis of bone collagen of extinct moa (Aves: Emeidae). Notornis 58: 1–7.
- 13. McCallum J, Hall S, Lissone I, Anderson J, Huynen L, et al. (2013) Highly informative ancient DNA ‘snippets’ for New Zealand moa. PloS ONE 8: e50732 doi:10.1371/journal.pone.0050732.
- 14. Cooper A, Poinar HN (2000) Ancient DNA: Do it right or not at all. Science 289: 1139.
- 15. Dickison MR (2007) The allometry of giant flightless birds. PhD thesis. Duke University.
- 16. Applegate TJ, Lilburn MS (2002) Growth of the femur and tibia of a commercial broiler line. Poultry Science 81, 1289–1294.
- 17. Davies SJJF (2003) ‘Rheas’.In Hutchins, Michael. Grzimek's Animal Life Encyclopedia 2nd edition, Farmington Hills, MI: Gale Group.
- 18. Bruning D (1970) Conjoined twin Darwin's rhea. The Wilson Bulletin 82: 219–220.
- 19. Reece RL, Butler R (1984) Some observations on the development of the long bones of ratite birds. Aust Vet J 61: 403–405.
- 20. Rawlence NJ, Metcalf JL, Wood JR, Worthy TH, Austin JJ, et al. (2012) The effect of climate and environmental change on the megafaunal moa of New Zealand in the absence of humans. Quarternary Science Reviews 50: 141–153.
- 21. Castanet J, Rogers KC, Cubo J, Boisard JJ (2000) Periosteal bone growth rates in extant ratites (ostriche and emu). Implications for assessing growth rates in dinosaurs. C R Acad Sci III 323: 543–550.
- 22. Tykot RH (2004) Stable isotopes and diet: You are what you eat. In Proceedings of the International School of Physics ‘Enrico Fermi’ Course CLIV, M. Martini, M. Milazzo, and M. Piacentini (Eds.) IOS Press, Amsterdam.
- 23. Wood JR (2007) Moa gizzard content analysis: further information on the diet of Dinornis robustus and Emeus crassus, and the first evidence for the diet of Pachyornis elephantopus (Aves: Dinornithiformes). Rec. Cant. Mus 21: 27–39.
- 24. Ambrose SH (1991) Effects of diet, climate and physiology on nitrogen isotope abundances in terrestrial foodwebs. J of Arch Sci 18: 293–317.
- 25. McMeans BC, Olin JA, Benz GW (2009) Stable-isotope comparisons between embryos and mothers of a placentatrophic shark species. J Fish Biol 75: 2464–2474.
- 26. Rogers KM (2009) Stable isotopes as a tool to differentiate eggs laid by caged, barn, free range, and organic hens. J Agric Food Chem 57: 4236–4242.
- 27. Hassan SM, Siam AA, Mady ME, Cartwright AL (2004) Incubation temperature for ostrich (Struthio camelus) eggs. Poultry Science 83: 495–499.
- 28. Roots C (2006) Flightless Birds. Greenwood Publishing Group, Inc., Westport, CT 06881.