Isolated island ecosystems are evolutionary microcosms typically exhibiting high levels of faunal endemicity (Mendelson and Shaw, 2005; Shaw and Gillespie, 2016). They are also often the last places to have been colonised by modern humans, as recently as hundreds to a few thousand years ago (Wilmshurst et al., 2011; Hansford et al., 2021), at times of relative climatic stability (Wanner et al., 2008; Waters et al., 2017). In island ecosystems modern human colonisation is frequently associated with widespread faunal extinctions and environmental modification (Perry et al., 2014; Hansford et al., 2021; Louys et al., 2021; Nogue et al., 2021).

Aotearoa New Zealand presents a unique opportunity to test for anthropogenic impacts on insular biodiversity. Its flora and fauna were shaped by tens of millions of years of isolation and regular long-distance dispersal (Wallis and Jorge, 2018), and exhibit a range of island adaptions including gigantism, flightlessness or flight-reduced terrestrial lifestyles, and slow K-selected breeding strategies (Worthy and Holdaway, 2002). At the time of East Polynesian colonisation in the late thirteenth to early fourteenth centuries CE (Wilmshurst et al., 2008, 2011), the New Zealand fauna was dominated by birds (>223 species), reptiles and marine mammals, specifically pinnipeds (four species of seals and sea lions). The avian fauna ranged from extinct birds such as nine species of giant flightless moa (Dinornithiformes; Bunce et al., 2009) to medium-sized ground-dwelling taxa including moa chicks or juveniles, giant goose (Cnemiornis spp.), adzebill (Aptornis spp.), moho (Porphyrio mantelli) and other rails, and several waterfowl (e.g., Finsch’s, blue-billed, musk and pink-eared ducks, merganser, and poūwa swan), as well as extant terrestrial birds including takahē (Porphyrio hochstetteri), kākāpō (Strigops habroptilus) and kiwi (Apteryx spp.; Worthy and Holdaway, 2002; Tennyson and Martinson, 2007). The top predators in the pre-human ecosystem were also birds including Haast’s eagle (Aquila moorei) and Eyles’ harrier (Circus eylesi; Tennyson and Martinson, 2007). Within reptiles, there were at least 110 species of Eugongylinae skinks and Diplodactylid geckos, seven Leiopelmatid frogs and one tuatara (Sphenodon punctatus) (Easton et al., 2017; Gemmell et al., 2020; Scarsbrook et al., 2021).

The arrival of Polynesians resulted in the widespread human-driven extinction of around 50% of the vertebrate biodiversity (Figure 1) as a result of hunting (Anderson, 1989; Holdaway et al., 2014; Perry et al., 2014) and anthropogenic environmental modification (McWethy et al., 2014), in addition to biological turnover events (Rawlence et al., 2017a), range-contractions (Salis et al., 2016), population bottlenecks (Rawlence et al., 2015a), and significant changes in indigenous forest cover through widespread anthropogenic burning (McWethy et al., 2014; McConnell et al., 2021). Within a few hundred years of human colonisation, the New Zealand ecosystem had been fundamentally transformed, which has left a clear signature in the archaeological and subfossil record (Holdaway and Worthy, 1996; Tennyson and Martinson, 2007; Holdaway et al., 2014; Perry et al., 2014; Walter et al., 2017; Waters et al., 2017; Rawlence et al., 2019). Polynesians also simultaneously introduced two exotic predators (Figure 1), the kiore (Pacific rat, Rattus exulans) and kurī (Polynesian dog, Canis familiaris) (Wilmshurst et al., 2008; Greig et al., 2018). Bones and teeth from both are found in colonisation-era archaeological sites throughout the three main islands, and on some but not all offshore islands (e.g., kurī have not been documented in archaeological sites on Rēkohu Chatham Islands ∼780 km east of mainland New Zealand but have been inferred, based on chewed bone, to have been present on the sub-Antarctic Auckland Islands ∼480 km south of the mainland; Davidson, 1987; Anderson, 2005; Greig and Walter, 2021). Well-documented archaeozoological analysis of early archaeological sites containing kurī bones also shows the presence of numerous extinct prey taxa (e.g., Anderson et al., 1996; Furey, 2002; Worthy and Holdaway, 2002; Scofield et al., 2003). Kurī were also a source of meat and industrial materials (e.g., kahu kurī dog skin cloaks, bone for tools and ornaments) for Māori, the direct descendants of those East Polynesian immigrants (Anderson, 1981, 1989; Davidson, 1987; Hartnup et al., 2011).

Much archaeological and palaeoecological research has focused on the direct impacts of humans (i.e., hunting, environmental modification; Anderson, 1989; Holdaway et al., 2014; Perry et al., 2014; Nogue et al., 2021) and predation by kiore - the latter based on modern ecological (Rayner et al., 2007; Ismar et al., 2014) and palaeoecological (Wilmshurst and Higham, 2004; Wilmshurst et al., 2008) studies. In contrast, the potential impacts of kurī in particular on the pre-European contact ecosystem have largely been overlooked as they are difficult to assess, despite the potential of dogs to be a major novel predator in recently colonised ecosystems (e.g., Koungoulos and Fillios, 2020; Hixon et al., 2021). Fleming (1962) stated “the simplest explanation is to attribute all late Holocene extinction to the profound ecological changes brought about by man with fire, rats and dogs,” while Anderson (1981, 1989) suggested kurī probably contributed to the extinction of moa. In contrast, Holdaway (1999) argued nearly all the known extinctions were due to people and kiore, and not kurī, which “probably had little effect on the biota.” Worthy and Holdaway (2002) took this idea one step further, boldly stating “the Polynesian dog can be exonerated: it was kept so close to camps that it is not a factor.”

Our aim is to discuss and reassess the potential ecological impacts of kurī on the pre-European New Zealand ecosystem, including their role in human-driven extinctions, and highlight future research directions that need to be addressed to fully understand the impact that dogs can have on insular island faunas.

An investigation of the impacts of kurī on Aotearoa New Zealand’s biota requires an understanding of kurī biology, behaviour and ecology. This is problematic, as by the mid-nineteenth century CE kurī were no longer identifiable as a distinct dog type in Aotearoa New Zealand, due to interbreeding and replacement by European dogs. Dogs were brought to New Zealand around the late thirteenth century CE by East Polynesian migrants, the end point of a major trajectory of human colonisation of the islands of the Pacific (Greig and Walter, 2021). Dogs were successfully transported and established on many islands of the region during these migrations. Molecular genetic studies show that Pacific dogs possess a distinctive mitochondrial genetic signature, distinguishable from other lineages, most likely originating from southern China (Oskarsson et al., 2012; Greig et al., 2018; Zhang et al., 2021). In the absence of extant populations, the bones, teeth and coprolites (i.e., desiccated faeces) of kurī recovered from archaeological sites now comprise a valuable source of information. Archaeozoological studies and emerging biomolecular techniques can provide data about kurī physical characteristics, diet, mobility, and genetic history. In addition to archaeological remains, there are numerous observations about kurī in early European historical literature, although most accounts focus on physical descriptions, sometimes with a brief comment about the uses of dogs by Māori (for a review see Colenso, 1877), rather than ecological information. In the absence of this type of information, modern studies of free-ranging dog populations and ethnographic data can assist with developing hypotheses for aspects of kurī behaviour.

Kurī feature in Māori mythology, oral histories and art forms, demonstrating their importance in Māori culture (Potts et al., 2013). Legendary kurī are associated with colonising voyages to New Zealand, historical events and the naming of landmarks. In everyday life kurī were kept as companions, watch dogs and hunting dogs, and as a source of meat and industrial materials (bones, teeth, and pelts) (Davidson, 1987). Written descriptions in early European accounts tell of a small, fox-like dog, with pricked ears and a bushy tail (Colenso, 1877). Morphometric analysis of skeletal remains indicates an adult shoulder height of slightly under 40 centimetres, and a body weight of 13–15 kilograms (Clark, 1997). This suggests a body shape similar to a small border collie, but more robust and with shorter legs. There is very little geographic and temporal variation in kurī skeletal remains, suggesting that there was no deliberate selection for particular morphological characteristics, such as size, that can be observed in the skeleton (Clark, 1997), despite their importance as a source of food and raw materials. Information about kurī life history, such as reproductive behaviour is extremely limited. Kurī are thought to have reached sexual maturity around 6 to 8 months of age, similar to modern dog breeds (Clark, 1995).

It appears that sizable kurī populations were able to be sustained in early Māori settlements (Figures 2A,B). The results of archaeozoological analysis of faunal assemblages from early Māori archaeological sites such as Wairau Bar, Shag River Mouth, Kaupokonui and Houhora, for example, have documented substantial numbers of dog bones (Anderson, 1981, 1989; Greig et al., 2018). Some later Māori settlements dating from the sixteenth and seventeenth centuries also appear to have supported high numbers of kurī, for example, Kohika (Horrocks et al., 2002, 2003) and the Masonic Tavern site (Wood et al., 2016). During Captain Cook’s second voyage to New Zealand, while anchored in Queen Charlotte Sound, both he and the ship’s scientist Forster commented that they saw plenty of dogs, including those travelling with people in canoes (Colenso, 1877).

Dogs are omnivorous generalists, capable of consuming and surviving on a wide variety of food types, ranging from human-derived garbage to animals that may be several times their body mass (Vanak and Gompper, 2009; Hughes and MacDonald, 2013). Kurī seen by early European travellers were not restrained or tethered (Clark, 1997) and had the freedom to roam throughout settlements and scavenge for food. This freedom would have enabled kurī to forage beyond the close confines of Māori villages (contra Worthy and Holdaway, 2002). Nineteenth century Māori villages often were observed with fenced enclosures or houses to prevent entry by dogs and pigs (the latter introduced by Europeans) (Earle, 1832; Dieffenbach, 1843). Many travellers described storehouses on poles or stilts several feet above the ground, used to protect food such as seed potatoes, or dried fish from roaming animals (Best, 1916). Left-over food was kept between meals in baskets on poles, for the same purpose (Earle, 1832). Stages or platforms were also documented to keep important objects away from kurī and kiore (Best, 1916).

Macroscopic analysis of kurī coprolites from archaeological sites suggests a varied diet consistent with the range of fauna commonly found in correspondingly dated Māori middens (rubbish heaps). Components include small bird, moa, fish, mollusc shell, as well as charcoal and other plant remains that often do not survive in middens (Clark, 1995, 1997; Irwin, 2004). A recent study of coprolites from the Masonic Tavern site in Auckland used both microscopic and ancient DNA methods, and identified fish, terrestrial and marine mollusc shell, charcoal, and wild and cultivated plant taxa (Wood et al., 2016). The contribution of faeces to the diet has also been noted for modern free-ranging dog populations (Butler et al., 2018). Preliminary isotopic analyses of kurī bones using carbon and nitrogen stable dietary isotopes further support a widely varied diet (Leach et al., 2003; Kinaston et al., 2013; Wood et al., 2017) incorporating terrestrial (i.e., similar isotopic compositions to herbivorous moa) and marine components.

To date, the only published evidence of kurī skeletal remains is associated with archaeological evidence for human activities (Figure 2B) rather than natural sites (e.g., caves, pitfalls, swamps, dunes), potentially creating a taphonomic sampling bias in available data. It is not clear whether kurī formed feral populations, although this is a possibility. There are few truly feral self-sustaining dog populations in the world today, where dogs live in a wild state independently of people. One example is in Australia, where dogs with a similar genetic ancestry to kurī were introduced around 4,000 years ago and subsequently became the dingo (Smith, 2015). Dingoes are an apex predator, and their prey includes a wide variety of Australian fauna and flora, including kangaroos, small mammals, birds, reptiles, fish, crabs, frogs, insects, and seeds, as well as introduced domestic species, including sheep (Smith, 2015). Kurī may have had a similar potential in New Zealand to form feral self-sustaining populations, particularly in warmer northern regions – indeed, recent feral dog populations have been observed in Northland (Piper, 2021) and “wild dogs” were observed subsisting on kākāpō, weka and ground dwelling birds in the nineteenth century (Anderson, 1981). The large proportion of dog bones in “natural” sites (i.e., no clear archaeological context) in northern New Zealand (Table 1) potentially supports this hypothesis (however, see Breadth and intensity of impacts below). Feral populations may have been at low abundance as seen with apex predators (Smith, 2015). Potts et al. (2013) raise the possibility that the monstrous or aggressive dogs in some Māori oral traditions may reflect concerns about kurī that had gone wild and beyond human control.

By the 1830s, with the introduction of European breeds to New Zealand, overall dog numbers appear to have increased to the point of being considered a nuisance. Dogs had been brought to New Zealand by Europeans as early as Captain Cook’s voyages in the late eighteenth century (Clark, 1995). Hunting dogs in particular were subsequently introduced relatively quickly by sealers and whalers (Wakefield, 1845). Earle (1832) considered dogs to be the worst introduction by Europeans, due to their rapid increase and the injuries they caused to other animals. Packs of wild dogs present in the mid-19th century appear to be European dogs (e.g., Thomson, 1859) which in the South Island coincide with the arrival of flocks of sheep in central regions. Dieffenbach (1843) observed that a native dog could not bring down a sheep (presumably because of its small size), but that cross-breeds and introduced dogs would do so. By the mid-1800s, kurī were no longer recognisable as a distinct breed, having been subsumed within the burgeoning European dog population (Clark, 1995) through interbreeding and genetic swamping.

Our review has identified a variety of potential impacts to indigenous fauna; direct predation, competition, habitat disturbance, pathogen transmission, and assistance with human hunting. We hypothesise that, given the close association between kurī and Māori, these impacts would have been widespread but relatively localised around Māori settlement or movement patterns (Figure 2B) (though we cannot discount the possibility of pre-European feral kurī populations), and would have significantly amplified the impact of humans on New Zealand’s biodiversity. Furthermore, unlike kiore, kurī impacts would have been focused on a sub-set of New Zealand’s avifauna, herpetofauna, and seals and sea lions, particularly mid-sized taxa (Figure 3). To obtain a full picture this impact, several avenues of future research need to be undertaken.

(1) Coprolites, attributed to “kurī” on the basis of size (and in some cases due to the presence of chewed bone fragments), have been found in numerous midden contexts. However, these coprolites could also be potentially attributed to other carnivores such as sea lions or omnivores like humans, especially given the dietary overlap between kurī and people (e.g., Horrocks et al., 2002, 2003; Wood et al., 2016). The use of ancient DNA is increasingly highlighting its utility as a genetic tool for the study of palaeodiet (e.g., Bon et al., 2012; Wood et al., 2020). Multidisciplinary analyses of genetically confirmed kurī coprolites (ancient DNA, macro- and micro-scopic, isotopic, palaeoproteomics) should be used to test hypotheses about geographical and temporal changes in kurī diet (and as a proxy for human diet), and how this reflects wider patterns in the previously recognised ecosystem change in New Zealand. These analyses should be conducted in conjunction with archaeozoological analysis of midden assemblages (including bulk bone metabarcoding of non-diagnostic material; e.g., Seersholm et al., 2018) and regional pre-human palaeofaunal surveys (e.g., Worthy, 1998b), especially of understudied taxa (e.g., small birds, herpetofauna), and modelling of Māori movement across the environment through time (e.g., is there a differential decline in ground-nesting birds?). Because kurī are closely associated with people, it may always be difficult to distinguish between hunting versus scavenging.

(2) Ancient DNA analysis of kurī bones and coprolites, and sedimentary archives from archaeological sites, focusing on the morphological and genetic identification of zoonoses (e.g., Irwin, 2004; Wood et al., 2013; Witt et al., 2021), could potentially shed light on this understudied aspect of faunal impact (e.g., did dogs bring zoonoses to Aotearoa, and were kurī a vector between seals, sea lions and people for tuberculosis?).

(3) To resolve the debate of whether there were feral pre-European kurī populations, and whether and when kurī had an independent impact of New Zealand’s biodiversity, multiproxy morphometric, genetic, radiocarbon and isotopic analysis of “natural” dog remains in sites with no clear archaeological context should be undertaken to determine if these are kurī or European dogs (or even hybrids), and feral or human-dependant dogs (e.g., are there differences in diet?) (e.g., Koungoulos, 2020; Runge et al., 2021; Witt et al., 2021; on dogs in other geographical contexts). Palaeontological excavations across New Zealand should also be aware of the potential for dogs in natural (i.e., non-cultural) subfossil sites. Integrated approaches incorporating palaeoenvironmental DNA techniques to detect ancient dog DNA in sedimentary archives should also be investigated (Rawlence et al., 2014; Dussex et al., 2021; Mitchell and Rawlence, 2021). It may not be possible to resolve this question but addressing the status of these dogs will allow more informed hypotheses to be made about the potential impact of feral kurī.

Kurī have largely been overlooked in contributing to the ecological consequences of Polynesian settlement of New Zealand as these impacts have been difficult to scientifically assess, with the majority of the research focusing on over-hunting, habitat destruction, and predation from kiore. Far from being “exonerated” we argue that kurī had the potential for a significant, widespread but relatively localised (cf. kiore) impact on New Zealand’s fauna. Given the behavioural characteristics of much of New Zealand’s avifauna (i.e., terrestrial, flight-reduced or flightless, ground nesting, slow breeding), they would have been highly vulnerable to predation by kurī. In this way, independent predation and as part of hunting trips, kurī mirrored and amplified the impact of people, especially we hypothesise for medium sized birds, herpetofauna, and seals and sea lions. Opportunities for future research will help resolve the remaining gaps in our knowledge of the impact of kurī on New Zealand biodiversity and on insular island ecosystems in general.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

KG and NJR devised the study, analysed the data, and wrote the manuscript. Both authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.