Detection of Salmonella during the translocation of two endemic New Zealand lizard species within the Hauraki Gulf

Abstract

Multiple species translocations to the same site are becoming common practice in New Zealand restoration programmes. With every new translocation, the risk of parasite transmission between populations can increase. The translocation of Hoplodactylus duvaucelii and Oligosoma smithi provided the opportunity to 1) test for Salmonella and Cryptosporidium occurrences; and 2) compare this Salmonella test prevalence with results from avian translocations at one release site, Tiritiri Matangi Island. Of the six reptile species tested, three skinks (O. aeneum, O. moco and O. smithi) and one gecko (Woodworthia maculatus) tested positive for Salmonella (n=274). This is the first record of Salmonella enterica subspecies IV 40:g,t:- in New Zealand reptiles. Test prevalences between reptiles and birds were similar, suggesting that Salmonella prevalence may be naturally low in the areas sampled.

Keywords:

• pathogen transmission
• New Zealand
• Motuora
• Korapuki
• Tawharanui Regional Park
• Hoplodactylus
• Oligosoma
• Woodworthia
• Sphenodon

Introduction

Translocation is an established conservation management technique in New Zealand (Sherley et al. 2010), but the intentional movement of populations between geographical locations poses concerns for potential pathogen transmission (Parker et al. 2006; Ewen et al. 2012a). The likelihood of transmission can increase with multiple translocations to one site, a practice common in many current restoration programmes (e.g. Gardner-Gee et al. 2007; Miskelly 2010). Due to such concerns, health screening is now required for translocations in New Zealand to minimise risk of exposure to certain pathogens (DOC 2004).

Health screens in reptiles often target generalist pathogens such as Salmonella and Cryptosporidium (Kaplan 1997; Warwick et al. 2001; Pasmans et al. 2008). Although Cryptosporidium infections are known to be lethal (Kaplan 1997; Pasmans et al. 2008), individual reptiles in which Salmonella are detected do not always express negative symptoms (Hoelzer & Wiedmann 2011). Therefore, concern centres on reptiles to potentially act as reservoirs for salmonellosis to other wildlife, agricultural/domestic animals and humans (Warwick et al. 2001; Clark et al. 2002; Chambers 2011). Unfortunately, there is a lack of baseline information on pathogen prevalence in wild populations (Parker et al. 2006; Ewen et al. 2012a). Such data are important for understanding the ‘natural’ prevalence in sympatric though diverse taxa and to provide meaningful interpretation to health screening results.

This study documents the test prevalences of Salmonella and Cryptosporidium in reptile populations at both source and release sites during the translocations of Duvaucel's gecko (Hoplodactylus duvaucelii) and shore skink (Oligosoma smithi) to two offshore islands and into captivity. We also compare our results to available avian Salmonella test results in the general context of pathogen transmission on Tiritiri Matangi Island.

Materials and methods

Two release sites, Tiritiri Matangi Island Scientific Reserve (220 ha) and Motuora Island Recreation Reserve (80 ha), and one mainland source site for O. smithi, Tawharanui Regional Park (558 ha), are located within the Hauraki Gulf. The source site for H. duvaucelii is Korapuki Island (18 ha) within the Mercury Islands Group, east of Coromandel. In November–December 2006, cloacal swabs (Sterilin) and, whenever possible, faecal samples were collected from resident reptiles at each release site and translocated animals during the quarantine period at Massey University (Auckland).

Isolations for Salmonella were conducted by inoculation of each swab samples into selenite cystine and tetrathionate broths for enrichment (Fort Richard, Auckland), incubated at 35–37 °C for 24 hours. Broths were then subcultured on xylose lysine desoxycholate and MacConkey plates from dehydrated media (Merck, Germany); incubated at 35–37 °C for 24 hours. Suspected Salmonella colonies were inoculated on triple sugar iron agar slants, urea slants and lysine decarboxylase tubes (Fort Richard, Auckland); incubated overnight at 35 °C. Further tests on the results, including agglutination, were conducted as required (Gartrell et al. 2007). Suspect colonies were sent to Environmental Science and Research for confirmation and serovar typing.

Faecal samples were screened using a commercially available indirect fluorescence assay (IFA) kit (Meridian Diagnostics Inc., USA) specific for Cryptosporidium oocysts, according to the manufacturer's instructions. An aliquot of each sample on a microscope slide was stained with IFA in a dark humidified chamber for 30 minutes and viewed by epifluorescent microscopy at 400× magnification.

Data on birds tested for Salmonella at Tiritiri Matangi Island between 2001 and 2008 were collated from various sources (Table 1). Due to low positive samples the true prevalence could not be calculated; therefore, the test prevalence (P_t) and 95% Wilson confidence intervals (CI; Reiczigel et al. 2011) were determined using Epitools (Sergeant 2009). Fisher's exact test (R v1.24; R Development Core Team 2008) was used to determine differences in Salmonella occurrences between species groups, reptile species and sites. For sites where all tests were negative, the maximum possible prevalence at 95% CI was calculated using WinEpiScope v2.0 (CLIVE, UK), under Detection of Disease. Despite limited data, the total numbers of reptiles on Tiritiri Matangi, Motuora and Korapuki Islands were estimated at 500,000 (Habgood 2003), 100,000 and 5000 individuals, respectively.

Table 1  Avian species tested for Salmonella on Tiritiri Matangi Island Scenic Reserve between 2001 and 2008, either as part of a translocation project or as general health screening. Data were collated from various sources, referenced below. Sample size, n, with the number of samples positive for Salmonella in parentheses.

Species	Year	n	Serotype
Bellbird (Anthornis melanura)	2001^1	4(0)	–
Fernbird (Bowdleria punctata)	2001, 2002^1	4(0)	–
Hihi (Notiomystis cincta)	2004^2	32(0)	–
	2005^2	77(0)	–
	2006^2	6(6)	Typhimurium DT195
	2007^3	73(0)	–
	2008^3	69(0)	–
North Island robin (Petroica australis)	2004^1	10(0)	–
	2007^1	21(0)	–
	2007^4	24(0)	–
Starling (Sturnus vulgaris)	2002^1	1(0)	–
Tomtit (Petroica macrocephala)	2004^1	6(0)	–
Tūī (Prosthemadera novaeseelandiae)	2001^1	5(0)	–
Whitehead (Mohua albicilla)	2004^1	8(0)	–
	2007^5	65(0)	–
	2008^5	91(0)	–
Total		496(6)
Notes: ^1 Parker et al. 2006; ^2 Ewen et al. 2007; ^3 T. Lovegrove, Auckland Council, unpublished data; ^4 Ewen et al. 2012b; ^5 K. Parker, Massey University, unpublished data.

Results

A total of 274 cloacal swabs and 28 faecal samples were collected from six reptile species, in which all individuals appeared in good condition when caught. Four samples tested positive for Salmonella, three were identified as Salmonela enterica subspecies IV 40:g,t:- (see Brenner et al. 2000 for serotype notation explanation) and the fourth was not able to be serotyped (Table 2). All faecal samples were negative for Cryptosporidium. We estimated that Korapuki Island could have a maximum prevalence of Salmonella presence of 0.42 that was undetected by our sampling effort. Higher maximum possible Cryptosporidium prevalences were estimated for Tiritiri Matangi (0.22), Motuora (0.53) and Korapuki Islands (0.22). No faecal samples were collected from Tawharanui.

Table 2  Overall Salmonella prevalence in six reptile species from Tiritiri Matangi Island Scientific Reserve, Motuora Island Recreational Reserve, Korapuki Island and Tawharanui Regional Park. Test prevalence P_t of Salmonella positives, with the 95% Wilson confidence intervals, in parentheses.

			Tiritiri Matangi Island		Motuora Island		Tawharanui Park		Korapuki Island
Species	n	Pt	n	Pt	n	Pt	n	Pt	n	Pt	Salmonella serovar
Oligosoma aeneum	34	0.03 (0.005–0.149)	13	0.00	21	0.05 (0.008–0.227)	–	–	–	–	Subspecies IV 40:g,t:-
O. moco	17	0.06 (0.01–0.270)	16	0.06 (0.011–0.283)	1	0.00	–	–	–	–	Subspecies IV 40:g,t:-
Woodworthia maculatus	30	0.03 (0.006–0.167)	30	0.03 (0.006–0.167)	–	–	–	–	–	–	Subspecies IV 40:g,t:-
Sphenodon punctatus	3	00.0	3	0.00	–	–	–	–	–	–	–
O. smithi	120	0.01 (0.001–0.046)	–	–	–	–	120	0.01 (0.001–0.046)	–	–	Salmonella untypable
H. duvaucelii	70	00.0	–	–	–	–	–	–	70	0.0	–
Total	274	0.02 (0.006–0.037)	62	0.03 (0.009–0.11)	22	0.05 (0.008–0.218)	120	0.01 (0.001–0.046)	70	0.0

There was no significant difference in Salmonella test prevalences between sites (P=0.214) or between species (P=0.195). The 496 samples from birds of Tiritiri Matangi had a Salmonella test prevalence of 0.01 (95% CI 0.06–0.026). No significant difference in Salmonella test prevalences was found between birds and reptiles on Tiritiri Matangi Island (P=0.175) or between Tiritiri Matangi Island birds and all reptiles (P=0.735).

Discussion

This study revealed the first record of S. subspecies IV 40:g,t:- in New Zealand reptiles. Salmonella subspecies IV is considered rarely pathogenic and is ubiquitous among reptiles or within the environment (Brenner et al. 2000). Indeed, the only other records for this serovar were from environmental samples and a domestic canine (ESR 2010). Due to its type, distribution, locality and low test prevalence, we suggest that S. subspecies IV 40:g,t:- is likely a natural occurrence on Tiritiri Matangi and Motuora Islands.

Although results were negative for Cryptosporidium, caution is needed in the interpretation of the outcome as test effectiveness in reptiles is unknown (Richter et al. 2011).

The overall Salmonella test prevalences from the four sites concur with previous New Zealand studies (Gartrell et al. 2007; Middleton et al. 2010; Ewen et al. 2012b) and are very low compared to other countries (Cambre et al. 1980; Parsons et al. 2010; Chambers 2011; Hoelzer & Wiedmann 2011). Therefore, it is unsurprising that the reptiles and birds had similar results. The New Zealand Public Health Surveillance reference laboratory has a 12% similarity in serovars detected between reptiles and birds in 1999–2008 (ESR 2010). Higher proportions of serovars are shared among wildlife and humans (100% with birds, 60% with reptiles). This suggests that humans may more frequently pose a risk to birds (and vice versa) as a possible source of Salmonella transmission compared to the risks associated with wild reptiles.

Some Salmonella test prevalences within species in our study contrast with results in Middleton et al. (2010), where positive samples were found in a smaller sample size of H. duvaucelii (P_t 0.095, n=21). This may be attributed to various reasons (Mitchell & Shane 2001), including different sampling and culture protocols between the studies. Furthermore, Salmonella detection using microbiological method is generally assumed to have 50% test sensitivity (test specificity of 98%; Middleton et al. 2010). The estimation of true prevalence for low-prevalence pathogens in wildlife populations such as in New Zealand will require a method with higher sensitivity value (Bager & Petersen 1991). There is a need for more accurate and standardised protocols for Salmonella detection to build an effective database.

Acknowledgements

We thank Manuela Barry, Dylan van Winkel, Chris Wedding, Graham Ussher, Melinda Habgood, Ben Barr and volunteers for field support, Hamish Mack and the late Jim Learmonth for laboratory work. Funding was received from Supporters of Tiritiri Matangi, Motuora Restoration Group, Massey University and Auckland Regional Council. Permits were supplied by Massey University (MUAE07/113) and Department of Conservation (DOCDM-73616).