A mid-Pliocene shearwater skull (Aves: Procellariidae: Puffinus) from the Taihape Mudstone, central North Island, New Zealand

Abstract

A fossil skull of Puffinus sp. (Aves: Procellariidae), 85 mm long, is reported from the mid-Pliocene Taihape Mudstone (3.220–3.110 Ma; Waipipian Stage) at Mataroa, near Taihape, central North Island, New Zealand. This is the first described evidence for Puffinus in New Zealand's pre-Pleistocene fossil record. The skull belonged to a shearwater of a species close in skull size and proportions to the present-day wedge-tailed shearwater (P. pacificus). It could be conspecific with P. pacificus, or of the ancestral species from which P. pacificus evolved. Certainty on that point is difficult with a sample of one and less than perfect preservation and exposure, however. The Mataroa fossil, and undescribed fossil evidence from Hawera, Taranaki, establish the presence of a shearwater fauna of at least several species in Pliocene coastal waters off the North Island of New Zealand. This is unsurprising given the importance of procellariiforms in New Zealand's current avifauna.

Keywords:

• fossil bird
• mid-Pliocene
• Procellariidae
• Procellariiformes
• Puffinus
• Taihape Mudstone

Introduction

The Tertiary fossil record of tube-nosed seabirds (Procellariiformes) was largely confined to the Northern Hemisphere until numerous South African seabird fossils were described in the 1980s (Olson 1985a , b). In the New Zealand region there is little pre-Pleistocene fossil evidence for procellariiforms despite their abundance in the modern avifauna and despite widespread marine fossil deposits that have yielded a diverse fossil record of penguins (Sphenisciformes; Fordyce 1991). Manu antiquus Marples, 1946, based on a fragmentary furcula from the Oligocene at Duntroon, north Otago, was originally thought to be an albatross (Diomedeidae). That has been ruled out (Olson 1985c), however, and the taxonomic affinities of this fossil are unknown.

A partial ulna attributed to a diving petrel Pelecanoides sp. was reported from the mid-Miocene Double Corner Shell Beds of North Canterbury (Scofield et al. 2006). Worthy et al. (2007) described Pelecanoides miokuaka from the upper Lower Miocene St Bathans Fauna (20.0–16.5 Ma) in Otago. However, Pliocene formations in New Zealand have produced few fossil vertebrates other than cetaceans and penguins (Worthy & Holdaway 2002). Undescribed material of two different shearwaters has been reported from Pliocene formations at Hawera, Taranaki, one of them said to be a Puffinus about the size of modern P. creatopus (Society of Avian Paleontology and Evolution Newsletter 6 [1992] and 7 [1993]; McKee 1994).

The present contribution to the documentation of the sparse pre-Pleistocene fossil record of New Zealand birds concerns a skull from the mid-Pliocene Taihape Mudstone at Mataroa, near Taihape.

Study site

On 22 December 2001, NH found the fossil skull at a road-cutting on Ridge Road North (grid reference 27433E 61663N; NZMS 260, T21, 1:50,000), 3 km south of Mataroa, in the central North Island west of Taihape. (The site has been registered as T21/f94 in the New Zealand archival database of fossil localities, ‘FRED’, of the Geological Society of New Zealand.) The road-cutting was c. 4 m high and the fossil was embedded in a grapefruit-sized spherical mudstone concretion lying separated from the bank from which it had eroded.

Given the site, the fossil is likely to have come from the Taihape Mudstone (PJJ Kamp, pers. comm. 2006; Waipipian Stage) which correlates with the Tangahoe Mudstone exposed in the west coast of the North Island south of Hawera. The fossil is probably from sediments that accumulated during the Gauss Chron and, in particular, within the normal subchron (3.220–3.110 Ma) lying between the Mammoth and Kaena Subchrons (Turner et al. 2005; PJJ Kamp, pers. comm. 2010). Other fossils found in concretions at the same site included cetacean bones and teeth, a large shark-tooth, vertebrae of bony fishes and molluscs.

Material and methods

The bird skull was partially embedded in grey cemented fine mudstone matrix. The cranium was whole but the nasal and premaxillary areas were recovered in seven main fragments that were glued together. BJG identified and compared the fossil by reference to bird bones in the collections of Auckland War Memorial Museum (AIM) and the Museum of New Zealand Te Papa Tongarewa (NMNZ), Wellington.

Measurements of the fossil skull and comparative material are listed in Table 1, and include: total length, the same character as CR in Fig. 1 of Holdaway & Worthy (1994); rostrum length from frontonasal hinge; cranium length from frontonasal hinge; and cranium width, the maximum width of cranium immediately posterior to the post-orbital processes.

Fig. 1  Diagram of reference points (A–D) used to assess proportions of the anterior end of the skull of shearwaters. See text for full explanation.

Table 1  Measurements (mm) of the Mataroa fossil shearwater skull (LB12868) and of the skulls of samples of modern shearwaters (Puffinus spp.). See Methods for definitions of characters (L: length). Mean sample size (in brackets) and range are given in that order. See Fig. 1 for reference points A–D. Bold typeface indicates that the range for the modern species overlaps the size (or ratio) of the fossil.

Character	LB12868	P. griseus	P. bulleri	P. pacificus	P. tenuirostris	P. gavia
Total L (TL)	87	92.5 (8)	89.5 (9)	87.0 (8)	79.5 (8)	76.6 (5)
		87–96	87–92	84–92	74–82	75–79
Rostrum L (RL)	48	50.7 (6)	49.1 (9)	48.0 (3)	41.0 (4)	42.0 (5)
		48–53	47–51	47–51	40–42	40–45
Cranium L (CL)	39	44.3 (6)	42.6 (9)	40.7 (3)	40.3 (4)	35.5 (5)
		44–45	42–43	40–42	40–41	35–36
Ratio RL:CL	1.18	1.15 (6)	1.15 (9)	1.18 (3)	1.02 (4)	1.19 (5)
		1.07–1.21	1.11–1.19	1.16–1.21	1.00–1.03	1.09–1.24
Cranium W	24	25.9 (8)	24.1 (9)	24.0 (8)	24.0 (10)	22.0 (5)
		24–27	23–25	23–26	23–25	22–23
Distance A–C	43	43.9 (8)	44.5 (9)	42.7 (8)	35.0 (8)	36.9 (5)
		41–45	44–46	40–45	31–38	35–39
Distance B–C (BC)	29	29.4 (8)	31.2 (9)	30.6 (8)	21.7 (8)	24.8 (5)
		27–31	30–32	28–33	19–23	23–28
Ratio BC:TL	0.33	0.32 (8)	0.35 (9)	0.35 (8)	0.27 (8)	0.32 (5)
		0.30–0.33	0.34–0.36	0.33–0.37	0.25–0.28	0.30–0.35
Distance B–D	13	14.7 (8)	10.9 (9)	12.4 (8)	12.0 (10)	12.0 (5)
		13–16	9–12	11–14	10–13	11–13
Width at A	12	13.9 (8)	13.2 (9)	13.6 (8)	12.6 (9)	10.8 (5)
		13–14	12–15	12–15	12–15	10–12
Width at B	7	7.1 (8)	7.4 (9)	7.9 (8)	6.5 (8)	5.2 (5)
		7–8	7–8	7–9	6–7	5–6
Depth at B	4	4.6 (8)	5.7 (9)	5.1 (8)	3.4 (8)	3.3 (5)
		4–5	5–6	4–6	3–4	3–4

To assess proportions of the anterior end of the skull, the following four reference points were identified (Fig. 1): A: narrow end-point of >-shaped notch in maxillary (end-point of notch is towards bill-tip) slightly ventral to D; B: distal end of external naris, as seen in dorsal view looking directly down on skull, i.e. distal end of hole that dorso-ventrally perforates nasal and underlying bones; C: distal tip of rostrum (premaxillary); and D: proximal end of external naris. The following measurements between or at these points were made: distance A–C (chord), distance B–C (chord), distance B–D, width at A, width at B, and depth at B.

Skulls of all genera of New Zealand marine birds in the avian osteological collection at AIM were used to confirm the identity of the fossil. In order to compile the measurements listed in Table 1, a number of shearwater skull specimens were measured (Table 2).

Table 2  Shearwater skulls measured for Table 1 (listed by museum registration number).

	Specimen
	AIM	NMNZ
Puffinus griseus	LB324, LB2147, LB10377, LB13130, LB13674, LB13889	18412, 24174, 25173–4, 25201, 26714, 26833
Puffinus bulleri	LB138, LB161, LB449, LB5800, LB9019, LB9254, LB9783, LB10940, LB13671	–
Puffinus pacificus	LB429–30, LB9235	11414, 16208–9, 21455, 23506
Puffinus tenuirostris	LB167–8, LB1347, LB12350	14992, 15908, 25134, 25557
Puffinus gavia	LB157, LB160, LB612, LB1488, LB13506	–

To observe details obscured by sediment, the skull was subjected to a micro-computerised tomography (CT) scan at the Auckland Bioengineering Institute (University of Auckland) using a Skyscan-1172 scanner (Skyscan, Belgium). It was scanned through 180° (with rotation steps of 0.4°) at a resolution of 17 µm using an Al-Cu filter. An opportunity arose later to have the fossil scanned at lower resolution at Ascot Radiology (Philips Brilliance 64-channel multi-slice CT-scanner) to generate a three-dimensional image. The fossil was scanned at 0.25 mm slice intervals with 0.5 mm slice thicknesses.

The fossil

The fossil skull (Fig. 2–Fig. 3 ; reg. no. LB12868) is held in the land vertebrates collection of the Auckland War Memorial Museum. It comprises the cranium, rostrum, palatines and jugal bar on the right side. The skull is pale- to dark-brown in colour, mineralised and has been slightly distorted during fossilisation. The foramen magnum and palatal area were pushed a little off-centre to the left, and the whole skull is slightly twisted along its length. The preservation and exposure of the fossil are generally fair but many fine, sharp ridges and processes on the skull have been lost. Many features of the ventral surface (Fig. 2C) and orbital region (Fig. 2B), including the lacrymal-ethmoid area, are obscured by sediment but are revealed to some degree in the reconstruction (Fig. 3). The skull is generally slender and elongate, with a very narrow rostrum. Although the tip of the rostrum is missing, the premaxilla clearly ended in a pronounced hook (Fig. 2B).

Fig. 2  Fossil shearwater skull (Puffinus sp.) from Mataroa (AIM LB12868): (A) dorsal; (B) right lateral; and (C) ventral. Background squares are 10 mm x 10 mm.

Fig. 3  The fossil shearwater skull (Puffinus sp.) as a multi-planar 3D reconstruction derived from CT scans: (A) dorsal; (B) left lateral; and (C) ventral. Each view is slightly oblique. (Reconstruction by T. Sheehan, Ascot Radiology, Auckland.)

The fossil is about 85 mm in total length, but some 1–2 mm of the tip of the rostrum is missing and we estimate the total original length of the skull to be about 87 mm. To allow comparisons with other skulls, all measurements involving the bill tip of the fossil were increased by 2 mm, e.g. rostrum length: 46+2=48 mm. The least width of the interorbital bridge is 10 mm. Further measurements are given in Table 1.

The fossil skull is that of a procellariiform seabird (Order Procellariiformes) as shown by a combination of features including the extent of the hooked premaxilla and size and shape of the nasal apertures. Other possible identities such as penguin (Sphenisciformes), cormorant (Pelecaniformes) and gull (Charadriiformes) were ruled out by comparison with reference material. Salt gland impressions around the dorsal perimeter of the orbits are visible on the right side. The cranium has large cerebral eminences and enlarged temporal fossae; it seems quite shallow dorso-ventrally but this is probably a distortion of fossilisation. The area proximal to the nasal apertures is depressed (Fig. 2B, 3B), not arched as in Diomedeidae.

The premaxilla is extremely slender in both dorsal and lateral views, and ventrally the interior nares are narrow (Fig. 3C). There is no external sign on the fossil skull, nor in the reconstructions (Fig. 3A, C), of large fused lacrymals. High-resolution CT-scans (e.g. Fig. 4) show the ectethmoids but no sign of adjacent lacrymals as would be expected if the lacrymals were fused to the skull. Serial scans anterior to the orbit were inspected and showed no sign of ectethmoids-lacrymals extending to or beyond the quadratojugal.

Fig. 4  CT scan showing transverse section of the fossil shearwater skull (Puffinus sp.) just anterior to the orbit (L: left side; V: ventral side; RE: right ectethmoid; Q: quadratojugal). (Scan by S. Malak, Bioengineering Institute, University of Auckland.)

These characters are the condition in the closely-related genera Puffinus and Calonectris (shearwaters) and exclude many other genera of Procellariidae such as Daption, Pachyptila and Pterodroma (e.g. Kuroda 1954). Close general agreement in all aspects of overall skull size, shape and proportions also clearly identify the fossil as belonging to a shearwater. Calonectris is separated from Puffinus mainly on post-cranial characters (e.g. Olson 2008), which do not help in the present case. Calonectris is currently far less diverse than Puffinus, and absent from the New Zealand region except as a rare straggler. We therefore assume the fossil belongs to Puffinus rather than Calonectris.

Comparisons with modern Puffinus skulls show that in size the fossil belonged to a species larger than the fluttering shearwater P. gavia (see measurements in Table 1). It was also therefore larger than Hutton's shearwater P. huttoni, little shearwater P. assimilis and Scarlett's shearwater P. spelaeus, which are all smaller on average than P. gavia (Holdaway & Worthy 1994; Heather & Robertson 1996). At the other extreme, the fossil skull is smaller than skulls of flesh-footed shearwater P. carneipes.

For 11 of the 12 characters in Table 1, the Mataroa fossil overlaps in size and proportions with the ranges for 2–4 modern species of Puffinus from the New Zealand region. There is no overlap for cranium length but it is only 1 mm short of overlap with the ranges for two modern species. Overall, the fossil is closest to the wedge-tailed shearwater P. pacificus (fossil dimension or ratio overlaps for 11 of 12 characters in Table 1) and sooty shearwater P. griseus (10 points of agreement). The fossil is not so close to Buller's shearwater P. bulleri and the short-tailed shearwater P. tenuirostris (5–6 points of agreement each).

Discussion

Miocene to Pliocene marine deposits in France, the US (Maryland, Virginia, North Carolina and California) and South Africa have now yielded numerous fossils attributed to shearwaters (Howard 1978; Olson 1985b; Olson & Rasmussen 2001). Many new species of middle to late Tertiary Puffinus have been described (e.g. Howard 1978), all based on post-cranial elements. In other cases (e.g. Olson & Rasmussen 2001), many of the Pliocene fossils have been attributed to extant species of shearwaters. So many Puffinus fossils have been found from so many geographic and temporal zones that a world-wide revision of them is needed in order to make proper sense of each individual find (Olson 1985c; Olson & Rasmussen 2001).

Procellariiform seabirds are thought to have evolved very slowly with some lineages seemingly having changed little morphologically since the late Oligocene (Olson & Rasmussen 2001). Puffinus bones from the Lee Creek Mine (Olson & Rasmussen 2001) included some related to, or similar to the size of, various living Atlantic species and some attributed to a lineage apparently ancestral to P. tenuirostris of the Pacific. Puffinus species B of Olson (1985b) from South Africa was nearest in size and morphology to P. tenuirostris, although perhaps slightly smaller.

The Mataroa shearwater skull could be conspecific with a contemporary species, and P. pacificus is the modern shearwater of the New Zealand region with which the fossil skull agrees most closely in size and proportions. P. pacificus is a warm-water species (Heather & Robertson 1996). New Zealand sea-surface temperatures declined during the Pliocene but were still no cooler than currently experienced in northern New Zealand (Stevens 1985), so the fossil skull could be expected to have belonged to a warm-water shearwater. However, with many details of the skull lost, distorted or not yet revealed by preparation, and with a sample-size of one, we are reluctant to assign it outright to P. pacificus.

Alternatively, it could represent a new species although we cannot simply assume from its age that it must be different from any species alive today. It is also possible that it belonged to one of the middle to late Tertiary species described from elsewhere in the world, the range of any of which might have extended to the New Zealand region. Most fossil Tertiary shearwaters are not known from the skull, precluding comparisons with the Mataroa fossil. For all these reasons we have not proposed a species-level identity for the Mataroa skull, and leave it as Puffinus sp. pending reconsideration in light of future studies.

The site of the shearwater skull at Mataroa, and the Pliocene site at Hawera, Taranaki, where material attributed provisionally to Puffinus was said to be found (Society of Avian Paleontology and Evolution Newsletter 6 [1992], and 7 [1993]; McKee 1994), are only 120 km apart. One of the Hawera shearwaters, reported as being about the size of modern P. creatopus (and therefore a large species), is too big to be conspecific with the Mataroa skull. Together these fossils establish the presence of a shearwater fauna in Pliocene coastal waters of the North Island of New Zealand, something to be expected for a region in which procellariiforms are today such an important avifaunal element.

Both fossil sites were in the seaway that separated the northern part of New Zealand (upper North Island) from the united southern section (lower North Island and South Island) during the Pliocene (Bunce et al. 2009: fig. S3). The fossils may represent birds that died at sea over the continental shelf, rather than birds that died on or near land and were washed into shallow marine waters.

Acknowledgements

We thank Alan Tennyson for facilitating access to the reference collection of seabird bones at NMNZ; Peter Kamp (University of Waikato) for advice on geological dating of the site; Iain Anderson and Sharif Malak (Bioengineering Institute, University of Auckland) and Tim Sheehan (Ascot Radiology, Auckland) for CT scans; Melanie Tuscia for help with the images; Jack Grant-Mackie for registering the fossil site with the Geological Society; and Trevor Worthy, Jack Grant-Mackie and two referees for helpful comments on the manuscript.