Historic Distribution

Trevor H. Worthy

Palaeofaunal Surveys, 2A Willow Park Dr., Masterton

Abstract

The distribution and relative frequency among anatids of Brown Teal (Anas chlorotis) in Holocene fossil deposits on the North, South, and Stewart Islands of New Zealand are reviewed. Brown teal remains representing 641 birds are reported from 73 widely distributed sites throughout New Zealand. The fossil sites indicate the prehistoric use of a large range of palaeohabitats including coastal sites, lakes, swamps, and forests as diverse as wet podocarp and beech forests up to 700 m altitude. There is often no direct association of these forest sites with aquatic habitats such as rivers or ponds indicating that Brown Teal were foraging often at considerable distances from such features while in forests.  In the seasonally drier eastern regions Brown Teal were confined more to aquatic habitats, though this may have been the result of competitive exclusion with Finschs duck which dominated terrestrial habitats, rather than habitat incompatibility.

Introduction

The Brown Teal (Anas chlorotis) is a small duck endemic to the main islands (North, South, Stewart, and associated islets) and the Chatham Islands of New Zealand. It is recognised as a distinct species as originally described following Oliver (1955), Kennedy & Spencer (2000), and Holdaway et al. (2001). Its nearest relatives are flightless congeners on Auckland and Campbell Island.

Brown teal were abundant historically throughout the North and South islands. It was common in kahikatea (Dacrycarpus dacrydioides) forest swamps and most commonly found by day sheltering beneath overhanging vegetation in quiet waterways, and was known to feed by night in a variety of habitats including lakes and in drains e.g. Buller, in Turbott (1967). However, within 40 years of Buller’s 1888 account, Brown Teal were noticeably declining in numbers, a trend that was related to loss of habitat (Oliver 1955, Turbott 1967). The wide variety of habitats used is summarised by Marchant & Higgins (1990), but includes, estuaries, tidal flats, beaches, rivers, hill swamps, gully-heads, mountain lakes, pasture, and forest.

The decline in numbers of Brown Teal has continued until the present day (Oliver 1955, and references in Williams 2001), a decline described most recently as ‘arguably, more dramatic than any other of New Zealand’s endemic birds’ (Williams 2001). At present (2002), the species exists in the South Island as a very few birds in Fiordland, and in the North Island by small and declining populations in Northland. Even the population on the apparent stronghold of Great Barrier Island is declining, with all population in imminent danger of extinction.

The aim of the present report is to clarify the prehistoric distribution of Brown Teal and the broad habitat (plant/landscape) characters with which remains have been associated throughout New Zealand, through an examination of fossil evidence. From these data I comment on the habitat preferred by Brown Teal prehistorically.

Methods

The fossil distribution is only assessed for the main islands of New Zealand (North, South and Stewart islands): records from Chatham Island are not assessed. Figure 1 shows the distribution of fossil sites, which were assessed for the presence of Brown Teal.

Data on the frequency and distribution of Brown Teal and other anatids were extracted from a database maintained by Palaeofaunal Surveys. Constituent data are derived from all published literature, particularly Worthy (1997a, b, 1998a, b, c, d); Worthy & Holdaway (1993, 1994, 1995, 1996, 2000); Worthy & Swabey (in press), Worthy et al. (in press) and unpublished theses and reports, particularly Millener (1981) and McGovern-Wilson (1986). All major sources are listed in the references. Archaeological data are based on Worthy (1999) and all the references cited therein.

Data from all sites (Appendix 1) are compiled into regional categories for the following selected anatids to show their relative abundance as fossils in each region (Table 1): Finsch’s duck (Chenonetta finschi), formerly in the endemic genus Euryanas, but recently referred to the Australian wood duck genus Chenonetta (Worthy & Olson 2002), grey teal (Anas gracilis), grey duck (Anas superciliosa), blue duck (Hymenolaimus malacorhynchos), scaup (Aythya novaeseelandiae), and merganser (Mergus sp.). The Auckland Island merganser (Mergus australis) was known historically only from the Auckland Islands. Fossil Mergus bones of Holocene age from Chatham Island have been suggested to be from a distinct taxon (Millener 1999), but no comparative study of the available material has been completed yet. Mergus bones of Holocene age from widespread sites in mainland New Zealand have not been studied, and so may be referable to M. australis, the Chatham form if it is indeed distinct, or a third distinct taxon.

In order to better associate fossil localities with palaeohabitats the analysis was restricted so far as practical to sites known or likely to be of Holocene age, as then palaeovegetation is much easier to reconstruct. Sites or faunas (if there are discrete faunas in a single site), known to be older than 10,000 years were not included.  However, as surface collected bones from some caves may include older material because time averaged deposits are common in New Zealand caves (Worthy & Swabey in press), undated individual bones/skeletons that are actually older than 10,000 years may be included.

Table 1. Frequency of selected anatids in fossil sites by region in New Zealand. Sites which are known to be exclusively older than Holocene in age are excluded. Data from Palaeofaunal Surveys database, and available on request.

Brown teal

Finsch’s duck

Grey teal

Grey duck

Blue duck

Scaup

Mergus

Totals

Northland dunes

12

4

5

4

0

4

1

27

Waitomo karst

50

36

0

2

6

3

0

94

Hawke’s Bay

7

14

0

2

1

3

1

26

Lake Poukawa Horn, 1983

421

46

338

355

0

165

2

1327

Inland Wairarapa

10

294

0

1

0

0

0

305

Coastal Wairarapa/Wellington

3

1

0

4

1

0

0

9

Takaka Hill & Valley

47

5

0

1

12

0

0

64

Mt Arthur/Owen

0

3

0

0

0

0

0

4

West Coast karst

4

2

0

0

2

0

0

8

Inland North Canterbury

16

100

3

4

0

5

0

128

Inland South Canterbury

9

133

0

0

3

0

0

145

North Otago

8

57

0

2

0

1

0

68

Central Otago

1

60

0

0

0

0

0

61

Southland

13

270

0

0

3

0

3

289

Coastal (Delaware /Marfells Bch)

27

3

2

15

1

12

10

70

Stewart Is (Masons Bay/Native Island)

13

0

0

0

0

0

9

22

Total

641

1028

348

390

29

193

26

2655

Notes to Table 1.

Brown teal have been recorded from 11 North Island, 16 South Island, and 1 (Old Neck) Stewart Island archaeological sites in addition to the natural fossil sites (Worthy 1999) (Figure 3).Grey teal (Anas gracilis) – In addition to the natural sites listed in Appendix 1, this species has been recorded in archaeological sites as follows: Taranaki (Kaupokonui, P21/3; Ohawe, N129/77); Coastal northern South Island (Marfells Beach P29/2; Haulashore Island, Nelson O27/56); North Canterbury (Redcliffs, S84/76; Whalers Bay Cave, O31/12); South Canterbury (Gooseneck Bend, H39/16, 1); North Otago (Pleasant River; Shag Mouth, J43/2); Southland (Riverton, D46/35; Tiwai Point, E47/13; Lee Island S131/4). Most if not all these records need to be treated with caution as the basis on which they were separated from Brown Teal was not given, and it is now known that the wing elements of both taxa widely overlap in size and are generally not distinguishable.

Merganser (Mergus sp.) – Additional records of this taxon are known from archaeological sites in the Hauraki Gulf (Ponui Island N43/1), Wellington coast (Paremata N160/50), Marfells Beach P29/2, North Otago (Kakanui J42/4, Shag Mouth J43/2), and Old Neck on Stewart Island.

Scaup (Aythya novaeseelandiae) – some of the Waitomo karst and Northland dune records require reassessment as Millener (1981) often confused scaup with Brown Teal (THW pers. observ.).

Blue duck (Hymenolaimus malacorhynchos) – Archaeological records for this species include at least one site in Central Otago (Hawksburn).

Distribution of Brown Teal

Brown teal bones representing at least 641 individuals are recorded from 73 fossil sites (Figure 1, 2).  The given frequency data is an under-estimate as data for three significant sites (Honeycomb Hill Cave complex, Megamania Cave, Poukawa N141/1-2) were not available. As these data are mainly based on literature records, there will likely be other unpublished records available from museum collections, but those given reveal the pattern of distribution and relative frequency of Brown Teal in relation to other anatids. The distribution of archaeological sites with Brown Teal remains is shown in Figure 3.

Northland dunes

The Northland dunes that have revealed most fossil deposits in the Far North of New Zealand include the dune fields of Doubtless Bay (Tokerau Beach) and others on the Karikari Peninsula, and those along the east coast of Aupouri Peninsula to North Cape and around to Cape Maria Van Diemen. All fossil deposits are in palaeosols developed under terrestrial vegetation – none are swamp or lake deposits. The data here are based on Millener (1981) who provided the latest summaries for each site. A considerable amount of more recent collecting by Brian Gill (Auckland Institute and Museum), Alan Tennyson (Museum of New Zealand Te Papa Tongarewa [MNZ]), and Worthy has occurred but summary data are not yet available. Anatids are a minor component of these faunas as shown by the total minimum number of individuals (MNI) of just 27. However, Brown Teal was the most abundant anatid species, and is likely to be underscored as the records of grey teal probably pertain to Brown Teal.

The palaeovegetation of the dunes in the Far North were reconstructed by Atkinson & Millener (1990) as a mosaic of coastal broadleaf/podocarp forest, coastal scrub, and dune grasslands.  Most dune sites were not directly associated with wetlands, though small streams flowed through some area, eg. the Tokerau Beach dunes, and swamps and small lakes backed onto some areas e.g. Cape Maria van Diemen. Atkinson & Millener (1990) listed Brown Teal and Finsch’s ducks as aquatic insectivores that were forest inhabitants. As Brown Teal was 3-4 times more abundant than typical aquatic ducks (grey duck, scaup) the suggestion that it was of more than incidental occurrence in these deposits, is supported.

Waitomo karst

The Waitomo karst is in the area between the Awakino and Waikato Rivers west of the Waipa River centred on Waitomo Caves in the west of the North Island. The karst landscape is developed mainly between the altitudes of 300 and 500 m asl. Anatids are relatively abundant in fossil sites in the Waitomo karst and Brown Teal is the most abundant species overall. All fossil faunas come from cave deposits, which are often time-averaged (Worthy & Swabey in press). This means, specimens of disparate ages are often grouped as single faunas. As Finsch’s duck was mainly a shrubland inhabitant, most of the Waitomo records for this species may be of Pleistocene age, and so the true frequency of Finsch’s duck in the Waitomo karst during the Holocene may be lower than indicated in Table 1.

One of the most important sites for Brown Teal fossils in the Waitomo area is F1c Cave (Worthy 1984).  There, Brown Teal were abundant in the upper Layers 2&4. The site is a pitfall trap adjacent to a small (10 x 30 m) flat depression in a broad valley floor. This depression was assumed by Worthy (1984) to have been a swampy and relatively open area that seasonally held water. While Carex sedges and treeferns were on the margins of this ‘wetland’, a podocarp (matai Prumnopitys taxifolia, rimu Dacrydium cupressinum) forest with associated hardwoods like tawa Beilschmiedia tawa formed a tall closed canopy forest all around the site. No streams flow on the surface in the valley around the site, and the nearest is 1-2 km distant.

While, the presence of Finsch’s duck in the Holocene is verified by its bones in the stratified sequence in F1c Cave (Worthy 1984; Worthy & Swabey in press) the data from F1c cave show changes in relative frequency of Brown Teal and Finsch’s duck over time. In the 1500-2000 14C yrs BP Layers 2&4, Brown Teal were three times as abundant as Finsch’s duck, whereas in the 12,000 14C yrs BP Layer 8, there were three Finsch’s duck and only one Brown Teal. This faunal change is paralleled by other changes in the compared avifaunas which Worthy & Swabey (in press) related to changes in the vegetation around the site. Regionally, the vegetation was dominated by shrubs in the late glacial, which were rapidly replaced by a closed forest in the earliest part of the Holocene. In the Pleistocene deposits of Gardners Gut Cave (Worthy & Swabey in press), Finsch’s duck dominates the anatid fauna. The presence of Finsch’s duck at F1c in the late Holocene may relate mainly to the presence of the small forest opening that was postulated to have been around the site.

Atkinson & Millener (1990) reconstructed the palaeovegetation of the Waitomo karst for most of the Holocene as a rimu/tawa forest. Podocarps, mainly rimu with some miro (Prumnopitys ferriginea), and northern rata (Metrosideros robusta) emerged over a tawa canopy. Hinau (Elaeocarpus dentatus) and mangeao (Litsea calicaris) were also common canopy components, with kamahi (Weinmannia racemosa) on steeper slopes. Shrublands and grasslands were absent or very rare in this area during the Holocene, and the only wetlands were small streams.  The incidence of Brown Teal fossils can be interpreted a lot more strongly than that they “may have used vegetated areas beyond stream courses” (Atkinson & Millener 1990). Most fossil sites in the Waitomo karst are of pitfall origin, with none known to have been accumulated by predators. As such, fossils have an origin as a live bird at the cave entrance, and as most sites are not stream submergences, and usually are 10s to 100s of metres from streams, then it must be presupposed that the birds were using forest near these entrances before they were entrapped.

Hawke’s Bay (excluding Lake Poukawa)

The sites grouped under the umbrella of Hawke’s Bay include coastal dune sites at Ocean Beach and inland rockshelters. The coastal dune sites were probably formerly vegetated in coastal broadleaf forest dominated by kohekohe (Dysoxylum spectabile) and karaka (Corynocarpus laevigatus) and were not associated with wetlands. The faunas in the inland sites were accumulated by avian predators, so the teal could have been taken either on wetlands (streams) or in forest glades. There are very few pitfall sites in which to trap faunas, so a direct comparison with Waitomo is not possible.

 Lake Poukawa.

The fauna from the Holocene deposits of Lake Poukawa, Hawke’s Bay, includes the largest collection of anatid remains known from New Zealand. There are two main and several minor sites at this locality and the fauna of only one (N141/12) has been described (Horn 1983). In Table 1 the values for the named taxon are the total for Layers 1-3 from Horn (1983). However, Horn also listed shoveler (Anas rhynchotis) and blue-billed duck (Oxyura australis). Specimens attributed to the latter are all now considered to be other taxa but mainly scaup (Holdaway et al. 2000). Several bones of the large extinct anatid Scarlett’s pink-eared duck (Malacorhynchus scarletti) were unrecognised as such and were confused with grey duck (pers obs.).  Horn did not state how grey teal was separated from Brown Teal, but it was likely on size, and as lengths of wing bones (most common elements by far) overlap in their ranges considerably (Worthy & Holdaway 1994) the identity of all small Anas bones need reappraisal.

In the early 1980s, the entire collection from all Poukawa sites was identified by P. R. Millener, however summary data has not been generated from these identifications (some 20,000 specimens). Recent, reappraisals of some of the identifications show many problems that preclude an accurate assessment of the anatid fauna. Firstly, several rare taxa, have been either overlooked (e.g. Malacorhynchus scarletti), or infrequently recognised (e.g. Mergus, Biziura delautouri). Smaller anatids were generally not identified and were labelled only as ‘anatid sp.’, and where identities were given, much confusion between shoveler, Brown Teal, grey teal and scaup bones is apparent. The Poukawa collection is in urgent need of reappraisal so that the relative frequency of anatids in a prehuman lacustrine fauna can be quantified.

However, Brown Teal were a very abundant component of the Poukawa fauna. It remains to be seen whether grey teal and shoveler were as abundant as Horn’s data suggests. Throughout the Holocene, Lake Poukawa was a shallow lake surrounded in tall podocarp forest dominated by matai, with a few forest openings (Wilmshurst et al. 1997).

Inland Wairarapa

Only a single site has revealed Brown Teal bones in the inland karst ranges of the Wairarapa. Karst areas occur mainly in the Waewaepa and Puketoi Ranges near Makuri, and in the Ruakokoputuna-Haurangi Range area. Sites in these regions are mainly pitfall traps. There are two major sites each with thousands of bones: the Coonoor site on the Waewaepa Range is a pitfall trap excavated about 1914, and Martinborough #1 (Haurangi Range) which was discovered in 1901 and excavated at various times, notably 1920, 1956, and in the 1960s. While much material from each site remains to be identified and accessioned into the collections, the many hundreds of bones that have been catalogued probably fairly indicate the composition of the source faunas. Brown teal are only known from Coonoor, and while the fauna is undated, a site in the impure limestone of this area is likely to be geologically young  (a few thousands of years at most) and can be expected to be of Holocene age. The Holocene vegetation in this area was a closed-canopy podocarp forest in a fairly wet climate (c.2000 mm per annum).

The absence of Brown Teal from Martinborough #1 probably means the species was absent from the surrounds of this site as many hundreds of other individual birds were represented. The site is located on a ridge in the Haurangi Range, which has a much lower rainfall than the Waewaepa Range. Finsch’s duck was common in Martinborough #1 with Millener (1981) recording a minimum of 277 individuals. An extensive series of radiocarbon dates on the site (R. N., Holdaway pers. comm.) indicates the material accumulated only in the last 3000-4000 years, so we can assume the present climate and nearby vegetation is representative of that which surrounded the site during deposition of the fauna. Seasonal drought is common and the nearby forest is dominated by mountain beech (Nothofagus solandri) with a few podocarps such as totara (Podocarpus totara) likely to have been emergent.

Coastal Wairarapa/Wellington

I have separated out the sites in the coastal fringe of the Wairarapa- Wellington region as the moa fauna of this zone indicates a different environment to that immediately inland. There was a thin zone of dune grassland and coastal scrub that backed onto a broadleaf forest (karaka/kohekohe) along much of this coast. The fossil sites are all in the dunes or swamp deposits behind beach ridges. Fossil faunas are only just beginning to be developed from the sites in this zone, as in the past only moa bones were routinely collected. As a result, very small non-moa faunas are as yet available.

A diverse fauna from high dunes north of the Mataikona River, deposited under a coastal forest or scrub with no associated wetlands at about 100 m asl, includes Brown Teal. In contrast, at Te Kaukau Point a swamp deposit behind a raised beach ridge preserves a mixture of wetland and terrestrial species, including Brown Teal. Most recently, Brown Teal has been recorded from a coastal swamp deposit at Tora on the Wairarapa coast (A. Tennyson, pers. comm.). So it may be expected that Brown Teal used the coastal forest and swamps along the foot of the hill behind coastal terraces along the Wairarapa coastline.

Takaka Hill

All the Brown Teal listed for the Takaka region are from pitfall sites on Takaka Hill (Worthy & Holdaway 1994). None of these sites are near streams and the area was or is vegetated in a silver beech (Nothofagus menziesii) forest with some cedar (Libocedrus bidwilli) and rare podocarps such as totara (Podocarpus totara) contributing to the canopy. The most important site is Hobsons Tomo, which had a spectacular assemblage of small birds including many ducks. Radiocarbon dating of the main deposits (Worthy & Holdaway 1994) indicates deposition has occurred from the present day back in time to at least 14,000 14C yrs BP. The old date was on a Finsch’s duck bone, and it is assumed that all four Finsch’s duck individuals, none of which where visible on the surface, were of pre-Holocene age. But there were also three blue duck and 38 Brown Teal present, most if not all of which will be of Holocene age. There is no wetland of any sort near the site and the nearest is over 1 kilometre distant. The evidence of this and several similar sites with fewer birds show that Brown Teal and blue ducks were using areas far from streams on Takaka Hill.

Mt Arthur and Mt Owen, northwest Nelson

These two localities have been separated out because they provide samples from the subalpine zone above the treeline, an area otherwise rarely represented in fossil deposits. No Brown Teal are known from sites close to or above the treeline in New Zealand, but that Finsch’s duck did use these habitats is shown by its presence in at least three sites. This is not just a function of taphonomy as other similar sized birds such as kakapo (Strigops habroptilus), weka (Gallirallus australis) and kiwi (Apteryx spp.) are common in subalpine sites.

West Coast karst

The data for the West Coast in Table 1 does not include any individuals from the Honeycomb Hill Cave System in the Oparara River headwaters. Some 20 km of surveyed passages in this system contain at least 70 discrete fossil sites known to be up to 20,000 years old (Worthy 1993). Brown teal are known from some of these sites, but summary data are not available. Similarly, Megamania Cave in the Gunner River has many discrete fossil sites and I observed Brown Teal in several of these in 1998, though most were not collected.  Both these cave systems lie in a mixed podocarp/ beech forest in a high rainfall area (c. 3000 mm per annum) at low altitude. The main canopy trees are rimu and red beech (Nothofagus fusca) with kamahi and rata on higher or steeper slopes, and there is a dense understorey with abundant lianas (kiekie Freycinetia banksii and supplejack Rhipogonum scandens), and moss covers most surfaces.

The rest of the data from the West Coast are derived from studies in the Punakaiki karst, mainly between Charleston and Fox River (Worthy & Holdaway 1993). Anatids were rare in the combined fauna from this area, which may be partly due to the fact that rich pitfall faunas of birds smaller than moa were lacking. The Holocene vegetation is assumed to be much like that present in the area now and is very similar to that in the Gunner River and the Oparara, described above.

Coast of northern South Island

Around the coast of the northern South Island from Nelson to Marlborough there are a number of fossil sites in dunes. The most significant of these are Delaware Bay, Marfells Beach, and an eroding dune south of Mussel Point, just east of Marfells Beach. All have Brown Teal remains in them.  The Delaware Bay site is on a spit between the estuary and the ocean, so the presence of Brown Teal indicates probable use of the estuary.  The Marfells Beach site is similarly located on a spit seaward of an extensive shallow lagoon, while the site south of Mussel Point is on a coastal flat beside a rocky shore.  All three sites indicate the late Holocene use of coastal habitats by Brown Teal. The extensive data for the Marfells Beach site in particular (Worthy 1998c) shows that Brown Teal were the most numerous anatid in a very diverse fauna (Table 2). It was markedly more common than any of the more aquatic ducks, and only the paradise shelduck (Tadorna variegata) approached it in abundance, perhaps reflecting the presence of a large area of saltmarsh in which to graze.

Inland North Canterbury

Within the North Canterbury region I am including sites on Mt Cookson and others around Waikari described in Worthy & Holdaway (1995, 1996).  Holocene Hole is a pitfall trap on the Mt Cookson plateau and the surrounding area was vegetated in a mountain beech forest in the late Holocene, while this deposit accumulated.  No wetland is near this site, in which two Brown Teal were present. All the other records come from sites near Waikari. Three (Glencrieff, Pyramid Valley, and The Deans) are wetlands – a spring, a lake, and a swampy earth flow respectively.  The fourth site, Waikari Cave, is a pitfall (the only one in the area), but is within a few tens of metres of a stream.

The anatid fauna of Pyramid Valley lake (Table 2) is significant as it shows that, in this shallow pond that was surrounded by tall matai forest during the period of deposition, Brown Teal were the most common duck. The absence of an aquatic macroflora is probably the reason why black swans (Cygnus atratus) were absent. One of the significant aspects of recent reassessment of the fauna at Pyramid Valley was that it revealed that grey teal were present prehistorically (Worthy & Holdaway 1996; Holdaway & Worthy 1997) apparently living with Brown Teal. While the regional fauna appears to be dominated by Finsch’s duck, this is mainly due to the large numbers from the pitfall site of Waikari Cave.  Finsch’s duck was derived from the wood duck (Chenonetta jubata) of Australia and is inferred to have been a grazing duck of terrestrial habitats (Worthy & Olson 2002). The drier regions in the east of the South Island afforded far greater areas of open vegetation, for example, scrub and grassland on the extensive river beds and on the higher slopes of hills, or on poorer soils, on which a grazing duck could feed. The Waikari Cave sample shows that in the Waikari area, Finsch’s duck dominated anatid faunas away from wetlands: it is much less common than Brown Teal in Pyramid Valley.

Table 2. A comparison of the anatid faunas from Pyramid Valley and Marfells Beach.

Pyramid Valley Marfells Beach
Brown teal 11 39
Finsch’s duck 5 3
Paradise shelduck 5 33
Grey duck 3 8
Grey teal 3 2
Scaup 5 5
Scarletts pink-eared duck 7 1
Black swan 0 22
Blue duck 0 1
Merganser 0 2
Musk duck 0 1

Inland South Canterbury

The faunas for inland South Canterbury are derived from the limestone areas of the inland downlands (Worthy 1997a). The sites include a mixture of swamp springs, pitfall traps in caves, and many predator sites accumulated by laughing owls or falcons.  The area is gentle rolling hill country 200 – 400 m asl, and during the late Holocene was probably vegetated in podocarp (matai-dominant) forest, with river flats containing areas of shrubland and grassland.

Most Brown Teal records came from the rich pitfall trap of Kings Cave. This site is located on the side of a hill about 100 m distant from a small valley and stream. Three dates are available from the site which all show a late Holocene period of accumulation (Worthy 1997a). Of the 281 birds represented in the fauna, 100 are Finsch’s duck, 68 are kiwi, 20 weka (Gallirallus australis) and 38 are kakapo (Strigops habroptilus), demonstrating the overall terrestrial nature of this fauna. The only typically dabbling duck was grey duck represented by one modern bone, but there were at least 3 blue ducks and 7 Brown Teal, illustrating the propensity of these species to be found distant from waterways.

North Otago

The North Otago faunas are derived from the area south of the Waitaki River that lies towards the coast from Duntroon in the limestone downlands (Worthy 1998a). Here, Brown Teal occur in at least two of the several fossil bearing swamp deposits that have been discovered in the valleys between low limestone hills.  In some valleys there are limestone cliffs in which predator accumulations have been found and Brown Teal are in two of these sites. There is only a single pitfall site (Ngapara), which is located a few hundred metres from any stream and is not near any ponds.  The Ngapara site accounts for most of the Finsch’s duck individuals, indicating that in this region, as in inland Canterbury, Finsch’s duck dominates the anatid fauna of terrestrial ecosystems.

The rolling hills of this area were probably vegetated much the same as the inland South Canterbury region discussed above. However, in the broad swampy valleys that typically lie between the hills, a swamp forest probably dominated by kahikatea (Dacrycarpus dacrydioides) was present, though other wetland podocarps such as yellow silver pine (Lepidothamnus intermedius) may have been present.

Central Otago

The single record of Brown Teal for central Otago comes from Earnscleugh Cave near Alexandra (Worthy 1998a). The fauna in this site is derived from both a pitfall and a predator accumulation. The site is several hundred metres up the side of a slope from the nearest stream. This site trapped most of the individual Finsch’s duck represented in the combined fossil fauna of the region, again reaffirming the dominance of this duck in dry eastern terrestrial ecosystems.

Southland

In Southland, most of the Brown Teal bones come from pitfall sites in low limestone ridges east of Winton (Worthy 1998b). While no Brown Teal bones are known from the several fossiliferous swamps in the region, this is probably more due to inept collection of material than their absence, as virtually no birds other than moa have been obtained from such sites. The ‘site’ termed Forest Hill Tomos is a general name for an unknown number of tomos or vertical pitfalls at Forest Hills, from which all the separate faunas were amalgamated as one, by the time of their study in the Otago Museum in 1997 (Worthy 1998b). Castle Rocks is the only one of the Brown Teal-bearing faunas to have been dated: a large series of radiocarbon dates indicates deposition during the late Holocene (Holdaway pers comm.). However, as most of the other collections are from surface deposits, a Holocene age is probably applicable to all. During this period the limestone ridge east of Winton (wherein lie the Forest Hills and McKerchars Cave sites) was clothed in tall rimu dominated podocarp forest.

In the regional avifauna, Finsch’s duck dominate the anatid fauna as they do in other eastern areas with low rainfall, because of their abundance in some tomos.

Stewart Island

Natural fossil sites are rare on Stewart Island with the main locality being the dunes of Mason Bay. However, there seabirds dominate with few land or freshwater birds recovered to date (Worthy 1998f). The most important site otherwise is a dune site on Native Island which is accessible from the main Stewart Island by wading at low tide (Worthy 1998d). There is no freshwater on Native Island, so the ducks present must have either been using the vegetated island habitat or the shoreline, which is very sheltered being within Paterson Inlet. On Native Island, Brown Teal were associated with mergansers, grey ducks, paradise shelducks, and black swan, all known to use or have used marine habitats. Finsch’s duck is not recorded from Stewart Island.

Discussion

The interpretation of the fossil deposits rests on the assumption that the birds represented were living in the immediate environment of the fossil site at the time of their death.  In no site in New Zealand is there evidence that brown bones were washed a significant distance, and so we can assume that the birds lived in the habitats surrounding the site. Secondly, it is also assumed that the while presence of one or two individuals may represent a random event such as a storm blown bird, the presence of several birds suggests that a population was present in the area of the site. This further necessitates acceptance that the birds were at home in the surrounding environment.

In this review of the distribution of Brown Teal fossils, I have compiled data for 641 birds from 73 sites of mostly Holocene age.  Finsch’s duck excluded, Brown Teal were the most abundant anatid in all regions of New Zealand. Because of the difficulties of identifying the exact habitat that surrounded a fossil site during the period of deposition and as fossil sites, with few exceptions, do not have more than one or two individual Brown Teal, a tight correlation between habitat type and relative frequency of anatids cannot be made. However, some useful insights into the variety of habitats Brown Teal used were obtained from amalgamating faunas from relatively small geographic regions likely to have had similar vegetation and climatic characteristics.

The broad trends revealed from the regional analyses

Brown teal were the most abundant species in all lacustrine habitats for which fossils are available, so they were formerly common on lakes. The data, especially that from sites such as Lake Poukawa and Pyramid Valley, supports the historical observations that Brown Teal preferred deep and quiet waterways with overhanging vegetation and dense kahikatea swamps. However, Brown Teal used to occupy a very much broader range of habitats than lakes and rivers, a range that was wider than that of any other anatid in New Zealand. Such additional habitats included coastal dunes, lagoons, and swamps, inland forests as diverse as wet rimu forests and seasonally dry matai forests in low rainfall areas, to montane silver beech and dry mountain beech forests up to 800 m altitude. Use of such forests was not restricted to swamps or streams within them, as fossil sites at times kilometres from any wetland indicate use of forest habitats far removed from wetlands. Generally, within a region, and in all habitats that Brown Teal occur, there is no predilection for water evident in the data. Thus, the historical observations that suggested a preferred habitat of quiet waterways etc are doubtless the result of ease of observations from waterways compared to those made deep within forests.

The only places Brown Teal may have usually avoided were the subalpine zone above the tree line and perhaps also the driest forests of eastern regions such as on the Haurangi Range in the Wairarapa. In the dry eastern regions where mosaics of forest, grassland, and shrubland occurred widely during the Holocene, Finsch’s duck dominated terrestrial sites and was apparently very abundant. As such it may have competed with Brown Teal excluding it from the more terrestrial habitats in these regions.

That the decline in Brown Teal numbers was first noticed not long after the introduction of mustelids suggests that predation by these predators was at least partly responsible for the decline of Brown Teal. However, that the species went extinct on both Chatham Island and Stewart Island, and are declining on Great Barrier Island, all places where mustelids are absent, suggests they are not the sole cause. The main predators on Stewart and Chatham Island are cats (Felis catus), Norway rats (Rattus norvegicus), and ship rats (R. rattus) indicating that Brown Teal populations cannot tolerate predation by one or other, or a combination of, these predators.

The fossil data cannot be analysed with statistical rigour that might lead to significant correlations with habitat types on a site by site basis. However, when site data is amalgamated into regions, wherein constituent sites have similar climate and vegetation variables, patterns if present should be evident. For example, we then see that Finsch’s duck is a common and abundant species only in areas with mosaics of shrubland and forest – mainly the drier eastern regions. The dominance of Brown Teal relative to other anatids (exclusive of Finsch’s duck) in composite faunas from each of the regions for which data is available, shows that Brown Teal was at home in a wide variety of habitats. The principal conclusion from these data is that Brown Teal did not have a preference for any specific habitat. It was a habitat generalist that could use any forest or wetland habitat below about 800 m, with the exception perhaps being the dry forests in areas receiving less than about 1000 mm rain annually. Therefore, the decline in Brown Teal populations throughout the last 100 years cannot be linked solely to loss of habitat, but conversely any of a wide range of surviving forest and wetland habitats should be suitable for the survival of populations of Brown Teal if mammalian predation is removed.

SELECTED REFERENCES

Atkinson, I. A. E.; Millener, P. R., 1991. An ornithological glimpse into New Zealand’s pre-human past. Acta XX Congressus Internationalis Ornithologici 1: 129-192. Holdaway, R. N.; Worthy, T. H., 1997. A reappraisal of the Late Quaternary fossil vertebrates of Pyramid Valley Swamp, North Canterbury, New Zealand.  New Zealand Journal of Zoology 24: 69-121.

Holdaway, R. N.; Worthy, T. H.; Tennyson, A. J. T., 2001. A working list of breeding bird species of the New Zealand region at first human contact. New Zealand journal of zoology 28(2): 119-187.

Horn, P. L. 1983. Subfossil avian remains from Poukawa, Hawke’s Bay, and the first record of Oxyura australis (Blue-billed duck) from New Zealand. Journal of the Royal Society of New Zealand 13(1/2): 67-78.

Kennedy, M.; Spencer, H. G. 2000. Phylogeny, biogeography and taxonomy of Australasian teals. Auk 117: 154-163.

Marchant, S.; Higgins, P.J. (co-ordinators) 1990: Handbook of Australian, New Zealand & Antarctic birds, vol. 1 – Ratites to Ducks. Melbourne, Oxford University Press.

McGovern-Wilson, R., 1986. Small bird exploitation. An archaeological approach to the study of fowling in Southern New Zealand. MA Thesis, Dept of Anthropology, University of Otago.

Millener, P. R. 1981. The Quaternary avifauna of the North Island, New Zealand. Unpublished PhD thesis, Department of Geology, University of Auckland, Auckland, New Zealand. 897 p.

Millener, P. R., 1999. The history of the Chatham Islands’ bird fauna of the last 7000 years – a chronicle of change and extinction. Proceedings of the 4th International meeting of the Society of Avian Paleontology and Evolution (Washington, D.C., June 1996). Smithsonian Contributions to Paleobiology 89: 85-109.

Oliver, W. R. B., 1955. New Zealand birds. A.H. & A.W. Reed, Wellington. 661 p.

Turbott, E. G. (editor), 1967. Buller’s birds of New Zealand. Whitcoulls Publishers, Christchurch Sydney London.

Williams, M., 2001. Productivity and survival within 2 declining populations of Brown Teal (Anas chlorotis). Notornis 48: 187-195.

Wilmshurst, J. M.; McGlone, M. S.; Partridge, T.R., 1997. A late Holocene history of natural disturbance in lowland podocarp/hardwood forest, Hawke’s Bay, New Zealand. New Zealand journal of botany 35: 79-96.

Worthy, T. H., 1984a. Faunal and floral remains from F1, a cave near Waitomo.  Journal of the Royal Society of New Zealand: 14(4): 367-377.

Worthy, T. H., 1984b.  An extensive subfossil deposit in Gardner’s Gut Cave, Waitomo. New Zealand Speleological Bulletin 130: 257-262.

Worthy, T. H., 1985.  The subfossil fauna of Aurora – Te Ana-au Cave.  Pp 14-21 In Williams, P. W. Ed.  Aurora – Te Ana-au Cave, a survey and interpretation of scientific resources.

Worthy, T. H., 1993.  Fossils of Honeycomb Hill.  Museum of New Zealand Te Papa Tongarewa. 56 pp.

Worthy, T. H., 1997a. The Quaternary fossil fauna of South Canterbury, South Island, New Zealand. Journal of the Royal Society of New Zealand. 27(1): 67-162.

Worthy, T. H., 1997b. Fossil deposits in the Hodges Creek Cave System, on the northern foothills of Mt Arthur, Nelson, South Island, New Zealand. Notornis 44: 111-124.

Worthy, T. H., 1998a. Quaternary fossil faunas of Otago, South Island, New Zealand. Journal of the Royal Society of New Zealand 28(3): 421-521.

Worthy, T. H., 1998b. The Quaternary fossil avifauna of Southland, South Island, New Zealand. Journal of the Royal Society of New Zealand 28(4): 537-589.

Worthy, T. H., 1998c. A remarkable fossil and archaeological avifauna from Marfells Beach, Lake Grassmere, South Island, New Zealand. Records of the Canterbury Museum 12(1): 79-176.

Worthy, T. H., 1998d. Fossil avifaunas from Old Neck and Native Island, Stewart Island – Polynesian middens or natural sites? Records of the Canterbury Museum 12(2): 49-82.

Worthy, T. H., 1998e. Fossil deposits in Megamania Cave, Gunner River, South Island, New Zealand. Conservancy Advisory Notes No 195. Department of Conservation, Wellington.

Worthy, T. H., 1998f. Fossils indicate Pelecanoides georgicus had large colonies at Mason Bay, Stewart Island, New Zealand. Notornis 45(4): 229-246.

Worthy, T. H., 1999. What was on the menu – avian extinction in New Zealand. NZ Journal of Archaeology 19: 125-160.

Worthy, T. H., 2000. Two late-Glacial avifaunas from eastern North Island, New Zealand – Te Aute Swamp and Wheturau Quarry. Journal of the Royal Society of New Zealand 30(1): 1-26.

Worthy, T. H., 2002. The New Zealand musk duck (Biziura delautouri Forbes 1892). Notornis 49: XXX 

Worthy, T. H.; Holdaway, R. N., 1993. Quaternary fossil faunas from caves in the Punakaiki area, West Coast, South Island, New Zealand.  Journal of the Royal Society of New Zealand 23(3): 147 – 254.

Worthy, T. H.; Holdaway, R. N., 1994. Quaternary fossil faunas from caves in Takaka Valley and on Takaka Hill, northwest Nelson, South Island, New Zealand.  Journal of the Royal Society of New Zealand 24(3): 297-391.

Worthy, T. H.; Holdaway, R. N., 1995. Quaternary fossil faunas from caves on Mt Cookson, North Canterbury, South Island, New Zealand. Journal of the Royal Society of New Zealand 25(3): 333-370.

Worthy, T. H.; Holdaway, R. N., 1996. Quaternary fossil faunas, overlapping taphonomies, and palaeofaunal reconstruction in North Canterbury, South Island, New Zealand. Journal of the Royal Society of New Zealand 26(3): 275-361.

Worthy, T. H.; Holdaway, R. N. 2000. Terrestrial Fossil vertebrate faunas from inland Hawke’s Bay, North Island New Zealand. Part 1. Records of the Canterbury Museum 14: 89-154.

Worthy, T. H.; Holdaway R. N., in press. The lost world of the moa: Prehistoric life of New Zealand.  Indiana University Press, Indiana.

Worthy, T. H.; Holdaway, R. N.; Alloway, B. V.; Jones, J.; Winn, J.; Turner, D. In press. A rich Pleistocene – Holocene avifaunal sequence from Te Waka #1: terrestrial fossil vertebrate faunas from inland Hawke’s Bay, North Island, New Zealand. Part 2. Tuhinga, Records of the Museum of New Zealand Te papa Tongarewa. (Dec 2001).

Worthy, T. H.; Mildenhall, D. C., 1989.  A late Otiran-Holocene paleoenvironmental reconstruction based on cave excavations in northwest Nelson, New Zealand.  New Zealand Journal of Geology and Geophysics 32: 243-253.

Worthy, T. H.; Olson S. L. 2002. Relationships, adaptations, and habits of the extinct duck ‘Euryanas finschi. Notornis 49: XXX.

Worthy, T. H.; Swabey, S. E. J. in press. Avifaunal changes revealed in Pleistocene – Holocene deposits near Waitomo Caves, North Island, New Zealand. Journal of the Royal Society of New Zealand ****.