Evidence of lizard predation by New Zealand kingfishers (Todiramphus sanctus vagans) and potential implications for threatened species translocations

Abstract

Dietary analyses were carried out on four species of native birds on Tiritiri Matangi and Motuora Islands, New Zealand, between February 2007 and February 2008. We examined regurgitated pellets of kingfishers, moreporks and swamp harriers, and the faeces of pukeko. Remains of lizards were found in 88% and 43% of regurgitated kingfisher pellets on Tiritiri Matangi and Motuora, respectively. All remains belonged to one species, Oligosoma moco. No evidence of lizard predation was found for the other three bird species. Translocation of lizards to areas supporting a high abundance of kingfishers should take avian predation risk into account.

Keywords:

• dietary analysis
• Oligosoma
• skink
• Tiritiri Matangi
• Motuora
• New Zealand

Introduction

In New Zealand, studies addressing the impacts of predation on lizards have focused almost entirely on introduced mammalian predators (Whitaker 1973; McCallum 1986; Miskelly 1997; Jones et al. 2005; Hoare et al. 2007). This has been in response to significant population declines and extinctions of endemic herpetofauna caused by mammalian predation (Whitaker 1973, 1978; Worthy 1987). The effects of predation by native animals on lizards are poorly understood, although there are records of lizard predation by invertebrates (Lettink & Patrick 2006), fish (Anonymous 2003), reptiles (Whitaker 1994) and birds (Marples 1942; Fitzgerald et al. 1986).

Internationally, birds have been regarded the most important predators of reptiles (Martin & López 1996). In New Zealand, at least two species of introduced birds, Eurasian blackbird (Turdus merula merula; Bell 1996) and common starling (Sturnus vulgaris vulgaris; Thompson 2000), and 12 species of native birds have been recorded preying upon lizards. These include banded rail (Rallus philippensis; Whitaker 1968), black-fronted tern (Chlidonias albostriatus; O'Donnell & Hoare 2009), North Island fernbird (Bowdleria punctata vealeae; Ball & Parrish 2005), kakapo (Strigops habroptilus; Falla et al. 1979), long-tailed cuckoo (Eudynamys taitensis; Oliver 1955), morepork (Ninox novaeseelandiae; Ramsay & Watt 1971; Imboden 1975; Haw & Clout 1999), New Zealand falcon (Falco novaeseelandiae; Fox 1977; Worthy 1997), New Zealand kingfishers (Todiramphus sanctus vagans; Mead 1947; O'Donnell 1981), pukeko (Porphyrio melanotus melanotus; Muggeridge & Cottier 1931; Carroll 1966), swamp harrier (Circus approximans; Pierce & Maloney 1989), takahe (Porphyrio hochstetteri; Whitaker 1991) and western weka (Gallirallus australis australis; Coleman et al. 1983). Reports of avian predation on lizards are largely based on anecdotal accounts and the few studies that have explored the predator–prey relationships between birds and lizards in New Zealand have done so with limited success (Schneyer 2001; Jordan 2005).

On offshore islands, native bird populations are known to reach considerable densities and resident lizard populations may provide an especially important food resource for avian predators (Schoener & Schoener 1978). Such predator–prey relationships have typically evolved within natural communities where established prey populations are generally robust enough to withstand the pressures exerted by natural predators (Bodini 1991). However, populations translocated for conservation reasons may not be able to withstand such predation, as even if a small proportion of a bird's diet consists of lizards, this may be sufficient to impact negatively on a founder lizard population (Martín & López 1996). A lack of understanding of predator–prey dynamics between native species could have significant impacts on the outcome of threatened species translocations. For example, unforeseen predation by pukeko and swamp harriers appears to have limited the population expansion of brown teal (pateke, Anas chlorotis) translocated to Tiritiri Matangi Island (Armstrong 2006). Similarly, the failure of a translocated population of threatened shore plover (Thinornis novaeseelandiae) to establish on Motuora Island was attributed to unanticipated predation by morepork (Aikmen 1999). The introduction of threatened bird species to islands outside of their natural range (e.g. takahe on Tiritiri Matangi Island) may result in unanticipated predation of lizards and could lead to their extinction from the island if the lizard population were small (Atkinson 1990). In New Zealand, recognition of the significance of lizards in functioning ecosystems has been an important impetus for their translocation to islands undergoing restoration (Whitaker 1978; Payton et al. 2002). Clearly there is a need for research and insight into lizard predation by native birds in order to understand the impacts of such predations. Such knowledge is important for the success of further translocations of threatened species (Armstrong & McLean 1995; Towns 1999).

In recognition of the lack of quantitative data, this study aimed to estimate the rate of avian predation on native lizards, on two offshore island sanctuaries, and provide a basis on which informed management decisions can be made during threatened lizard species translocations. This study was conducted on two offshore islands; Tiritiri Matangi (36°36′ S, 174°53′ E, 220 ha) and Motuora (36°30′ S, 174°47′ E, 80 ha) Islands (Hauraki Gulf, Auckland, New Zealand) as part of a wider lizard translocation and monitoring project. Introduced mammals have never established on Motuora (Gardner-Gee et al. 2007); however, Tiritiri Matangi once supported populations of kiore (Rattus exulans), rabbits (Oryctolagus cuniculus), cats (Felis catus) and goats (Carpus hircus). By 1993, all introduced mammals had been eradicated from Tiritiri Matangi (Mitchell 1985; Galbraith & Hayson 1994). These islands are currently free of introduced mammals and are subject to ecological restoration by way of re-vegetation and faunal re-introduction programmes. These two islands share four species of lizards, while Tiritiri Matangi supports one further lizard species and tuatara (Sphenodon punctatus) (Table 1).

Table 1  Status and density of lizard species present on Motuora and Tiritiri Matangi Islands.

	Motuora		Tiritiri Matangi
Species	Status	Density	Status	Density
Oligosoma aeneum	Remnant	Not available	Remnant	6173/ha^a
Oligosoma moco	Remnant	Not available	Remnant	4617/ha^a
Oligosoma smithi	Translocated in 2006	30 individuals	Translocated in 2006	30 individuals
Hoplodactylus duvaucelii	Translocated in 2006	20 individuals	Translocated in 2006	19 individuals
Woodworthia maculatus	Absent	–	Remnant	Not available
^aDensity estimates were calculated from within for grassland habitats (Habgood 2003).

Materials and methods

Dietary analyses were conducted on four resident species of native birds known to prey upon lizards. These included pukeko, kingfisher, morepork and swamp harriers. Diets were assessed using several techniques, including analyses of regurgitated pellets, faecal material, nest material and stomach contents (where available).

The two islands were systematically searched for kingfisher and morepork nest and roost sites between February 2007 and February 2008. Faecal and regurgitated pellet samples were collected from across the islands and frozen to prevent decay and damage by invertebrates. Nest samples were collected from morepork and kingfishers after each successive fledging to avoid disturbing the nesting birds. Faecal samples from pukeko were collected from across the islands every month. Dead birds of these four species were collected opportunistically as they became available, and stored in a −20 °C freezer until examination. The entire alimentary tract (i.e. oesophagus, gizzard, stomach and intestines) of deceased birds was examined for the presence of lizard remains.

All regurgitated pellets were examined under a dissecting microscope. The prey composition from each pellet was expressed as the frequency of pellets containing a prey item. Prey items included the following categories: insects, decapods (e.g. crabs), fish and lizards. The number of individual lizards per pellet was estimated by counting the number of brain cases, jaw bones (i.e. two dentaries and two maxilla per lizard), frontal bones and pelvic girdles. Lizard remains were identified to species by morphological characteristics of the dentaries and maxilla (Worthy 1987), and by comparisons with museum reference skeletons. Prey remains in nest samples were not quantified since some remains may have been deposited during previous nesting seasons. Seasonal dietary analyses were not carried out because of the lack of comparable data across seasons.

Results

One hundred regurgitated pellets were collected from nine individual kingfishers on Tiritiri Matangi and seven pellets collected from three individual kingfishers on Motuora. Lizard remains were present in 88 (88%) of the Tiritiri Matangi kingfisher pellets and in three (43%) of the Motuora kingfisher pellets. Details of pellet contents are shown in Fig. 1. All lizard remains were identified as one species, Oligosoma moco, based on the size and shape of the jaw bones. The average number of lizards per pellet varied depending on the type of skeletal remains present in the pellets (i.e. braincase, jaw bones, frontal bone and pelvic girdle) (Table 2). Conservative estimates of the number of lizards per pellet were two per pellet on Tiritiri Matangi, and one per pellet on Motuora.

Figure 1  Percentage of regurgitated kingfisher pellets containing prey items from Tiritiri Matangi (n=100) and Motuora (n=7) Islands.

Table 2  Average number (±SEM) of skink bones recorded in regurgitated kingfisher pellets containing lizard remains, on Tiritiri Matangi and Motuora Islands in 2007.

	Lizard remains/pellet
Island	n ^a	Braincase	Jaw	Frontal bone	Pelvic girdle
Tiritiri Matangi	88	1.8 (±0.2) [0–6]	13.2 (±1.6) [0–29]	3.6 (±0.3) [0–8]	2.1 (±0.3) [0–7]
Motuora	3	0.5 (±0.5) [0–2]	0.7 (±0.7) [0–2]	0.5 (±0.5) [0–2]	0.3 (±0.3) [0–1]
Numbers in square brackets represent the range. ^aNumber of regurgitated kingfisher pellets containing lizard remains.

Two regurgitated pellets were collected from swamp harriers on Tiritiri Matangi, and neither of these pellets contained lizard remains. No swamp harrier pellets were collected on Motuora.

Lizard remains were found in six (60%; n=10) kingfisher nests on Tiritiri Matangi, but were recorded less often in nests on Motuora (23.8%, n=21). Remains of all other prey categories were detected in nest holes on both islands.

No lizard remains were identified in regurgitated morepork pellets from Tiritiri Matangi (n=67, 12 birds) or Motuora (n=107, nine birds). Prey consumption consisted almost entirely of invertebrates, with wētā (Orthoptera: Anostostmatidae) occurring in 85.1% and 89.5% of owl pellets on Tiritiri Matangi and Motuora, respectively. Bird remains were occasionally apparent in pellets from owls on both islands. No lizard remains were found in morepork nests on either island.

Pukeko faecal samples from Tiritiri Matangi (n=50) and Motuora (n=50) were comprised entirely of vegetative matter and did not contain lizard remains. Similarly, no lizard remains were observed within the guts of a swamp harrier and partially decomposed pukeko.

Discussion

Evidence of lizard predation was found only in the diet of kingfishers. The high proportion of kingfisher pellets and nests containing lizard remains, relative to the other identified prey categories, suggests that lizards form an important component of the kingfishers’ diet on these islands. Previous studies have also noted the importance of lizard prey items in the diet of both adult and nestling kingfishers (Burton 1998; Blomberg & Shine 2000; Asokan et al. 2009; Soud et al. 2010). For example, Asokan et al. (2009) reported that 24% of white-breasted kingfisher (Halcyon smyrnensis) regurgitated pellets contained the remains of lizard prey.

All lizards consumed by kingfishers in our study consisted exclusively of O. moco. This lizard is a diurnal heliotherm and is common on the islands, with an estimated density of up to 4617 per ha in grassland habitat on Tiritiri Matangi (Habgood 2003). The diurnal, sun-basking activity exhibited by this species is likely to have contributed to the high rate of predation by visually oriented predators such as kingfishers (Anholt & Werner 1975). The vulnerability of diurnal heliotherms to predation is emphasised by the lack of Oligosoma aeneum (a crepuscular species) in the diet, despite this species occurring at densities of up to 6173 per ha in grassland habitat on Tiritiri Matangi (Habgood 2003).

Predation is a function of prey availability (Anholt & Werner 1975), and the higher proportion of lizard remains in kingfisher pellets on Tiritiri Matangi, compared with that of invertebrate remains, may be related to high lizard prey abundance, or perhaps reflects selection of more protein-rich prey items, especially during periods of high energy demand such as breeding seasons (O'Donnell & Hoare 2009). However, the overall impact of kingfishers on the island lizard populations could not be determined because of the lack of information pertaining to demographics and structure of the lizard populations.

Despite the high density estimates of O. moco and O. aeneum on Tiritiri Matangi, lizard remains were not detected in the diet of morepork, pukeko and swamp harrier in this study, even though these birds have been recorded preying on lizards. This may indicate that lizard predation by these species is low and occurs opportunistically. However, larger sample sizes would be required to verify this.

Translocations

In New Zealand, lizard conservation efforts have focused on translocating threatened species to offshore islands free of introduced mammalian predators (Armstrong & McLean 1995; Towns & Ferreira 2001). Founder populations are generally small because of the limitations of harvesting source populations, and opportunistic predation of founders may be great enough to exacerbate the effects associated with K-strategist lizard species. For example, following the translocation of highly threatened Cyclodina whitakeri to Korapuki Island, population models suggested that the population became highly vulnerable to collapse if predation of adults exceeded 5% per annum (Towns 1994).

Translocating reptiles into unfamiliar environments has potential to disrupt their behaviours, especially as reptiles may move more frequently to explore, or as a result of disorientation (Plummer & Mills 2000; Sullivan 2004). Increased activity and frequency of movement through unfamiliar habitat may ultimately affect the survival and establishment of a lizard population because of increased risk of predation (Plummer & Mills 2000). Post-release monitoring of 30 translocated Oligosoma smithi on Motuora has revealed a decrease in detectability through time (M. Baling pers. comm. 2011) following their release at a beach site known to support kingfishers. Although the ultimate cause of this decline in delectability cannot be ascertained, a combination of storm damage to habitat and predation by kingfishers could be proximate causes. In comparison, the O. smithi population translocated to Tiritiri Matangi has shown a marked increase in numbers. The success of this population establishment has been attributed to the provision of adequate natural refuges throughout the release habitat, which are likely to have offered the skinks protection from storm events and avian predators. Future studies involving lizard population monitoring and kingfisher dietary analyses following the second release of O. smithi onto Motuora in February 2011 may provide a more rigorous insight into the effects of avian predation on small translocated lizard populations.

Assessing the risk of predation on translocated lizard populations may incorporate approaches such as modelling possible outcomes based on known life history traits of the lizard species involved. An understanding of predator–prey dynamics between birds and lizards, and the associated risks, can potentially improve the outcome of threatened species translocations. Conservation managers should be aware of the potential impacts of resident birds, notably kingfishers, on small native lizards and should consider these factors when planning translocations of threatened endemic lizards; especially diurnally active species. Strategies to mitigate predation risk include: avoiding the release of lizards at locations or microhabitats where densities of avian predators are known to be high, releasing lizards outside of the avian predator breeding season, selecting release sites with adequate natural refuges and food resources, and enhancing habitat with additional refuges (artificial or otherwise) (Webb & Shine 2000; Lettink 2007). In addition, removing potential hunting posts (e.g. disused fence posts, dead tree stumps) at the release sites might aid in reducing the risk of kingfisher predation of released lizards.

Acknowledgements

We are grateful to Richard Sharp and Helen Lindsay for their abseiling skills, to Dave Jenkins for his assistance on Motuora Island, and to Trent Bell, Sarah Wells and Marleen Baling for their constructive comments. Funding was provided by the Society for Research on Amphibians and Reptiles in New Zealand Herpetological Research Award, the Supporters of Tiritiri Matangi Research Fund, the Motuora Restoration Society and Massey University. This study was endorsed by the Department of Conservation (permit AK-19734-FAU) and Massey University Animal Ethics Committee (Protocol 06/98).