Author: Ifor L. Owen Columba Awui Eric Langelet Wenda Soctine Simon Reid

Title: The probable role of cannibalism in spreading T. papuae infection in a crocodile farm in Papua New Guinea Author: Ifor L. Owen Columba Awui Eric Langelet Wenda Soctine Simon Reid PII: S0304-4017(14)00217-9 DOI: http://dx.doi.org/doi:10.1016/j.vetpar.2014.04.008 Reference: VETPAR 7215 To appear in: Veterinary Parasitology Received date: 12-12-2013 Revised date: 3-3-2014 Accepted date: 6-4-2014 Please cite this article as: Owen, I.L., Awui, C., Langelet, E., Soctine, W., Reid, S.,The probable role of cannibalism in spreading T. papuae infection in a crocodile farm in Papua New Guinea, Veterinary Parasitology (2014), http://dx.doi.org/10.1016/j.vetpar.2014.04.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1 1 2 The probable role of cannibalism in spreading T. papuae infection in a crocodile farm in Papua New Guinea 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Ifor L. Owen a, Columba Awui a, Eric Langelet b, Wenda Soctine b, Simon Reid c * a National Veterinary Laboratory, National Agriculture Quarantine & Inspection Authority (NAQIA), P O Box 741, Port Moresby, Papua New Guinea b Crocodile Farm, Mainland Holdings Ltd., PO Box 169, Lae, Morobe Province, Papua New Guinea School of Population Health, the University of Queensland, Herston Road, Herston, QLD, 4006, Australia *Corresponding author e mail address: simon.reid@uq.edu.au Abstract Between 2003 and 2007, 83 (50%) of 167 crocodiles (Crocodylus porosus) purchased as juveniles by a crocodile farm 3 or 4 years earlier from Kikori, Gulf Province, were found to be infected with Trichinella papuae. Between 2005 and 2007 infection was detected in a number of crocodiles at the farm obtained from six localities other than Kikori, as well as in a few animals born on the farm. Up to 2004, all juveniles at the farm, whether wild or farmborn, were penned together; the practice was then stopped to prevent possible infection through cannibalism. The last infected animal from Kikori was seen in 2007, 4 years after the purchase of crocodiles from there ceased. The last non Kikori infected crocodile was seen, also, in 2007. None of the 1,972 crocodiles (comprising wild and farmborn animals) tested from 2008 to 2013, using the digestion method, was infected with T. papuae. This indicates that infection of non Kikori crocodiles was the result of cannibalism within the farm during the years up to 2004 when juvenile crocodiles were kept together, and that the farm is now free of the infection. Keywords: Trichinella papuae, Crocodylus porosus, cannibalism, crocodile farm management 23 24 25 26 1. Introduction Trichinella was known only from warm blooded animals until a new non encapsulated species, T. zimbabwensis was reported in crocodiles (Crocodylus niloticus) in Zimbabwe in 1995(Pozio et al. 2002). Another non encapsulated species, T. papuae, first discovered in pigs in Papua New Guinea (PNG) (Pozio Page 1 of 11

2 27 28 et al. 1999), was later (2003) found infecting saltwater crocodiles (Crocodylus porosus) from one locality in PNG, Kikori, (Pozio et al. 2004). It has been implicated in cases of trichinellosis in humans in Thailand 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 (Khumjui et al., 2008; Kusotsuk et al., 2010), and there is serological evidence of human infection in PNG (Owen et al., 2005). Trichinella cycles in nature through host scavenging or cannibalism. Prior to the discovery of infection in crocodiles, recycled raw crocodile meat was known to be used as crocodile feed, a habit that could potentially perpetuate infection in crocodile farms (Mukaratirwa & Foggin, 1999). In PNG, saltwater crocodiles were naturally infected with T. papuae as a result of feeding young captured crocodiles with infected wild pig meat while they were penned before sale to a crocodile farm (Pozio et al., 2005). PNG has one large commercial crocodile farm that holds about 30,000 C. porosus obtained through its own breeding programme and by purchasing eggs and young crocodiles from rural villagers under a Sustainable Management Scheme. The young, wild born crocodiles measure about 45cm when bought and require 3 to 4 years at the farm to attain a size suitable for sale of skins. As the farm exports meat for human consumption as well as skins, quarantine authorities require testing of meat for Trichinella. This began in late 2002. The crocodile farm bought wild born animals from about 31 localities in lowland PNG between 2002 and 2012, although purchases were predominantly from Ambunti, Angoram and Wewak (East Sepik Province), Popondetta (Oro Province), Kimbe (West New Britain Province) and Kikori (Gulf Province) (Fig. 1). The farm ceased buying young crocodiles from Kikori in 2003, as soon as animals from there were 46 found to be infected with T. papuae. 47 48 49 In 2005 T. papuae infection was detected in farm born as well as crocodiles from locations other than Kikori held at the farm. The discovery prompted an investigation to determine the source and mode of transmission. Page 2 of 11

3 50 51 2. Materials and methods Quarantine protocols required testing of meat from 10% of animals at each quarterly cull. Samples 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 (2.5g each) of masseter, intercostal, biceps and tail muscle were pooled (10g) from individual crocodiles and tested separately, using the standard HCl/pepsin digestion method (Gamble et al., 2000). The work was carried out by trained personnel at the National Veterinary Laboratory, National Agriculture Quarantine and Inspection Authority (NAQIA), PNG. The proportions of wild born and farm born crocodiles, and the proportions of male and female wildborn crocodiles that were infected with Trichinella and the 95% binomial confidence intervals were determined. The statistical significance of the difference between the prevalence of infection in male and female crocodiles was determined using the Chi square test with a 95% confidence limit. The probability that the farm born crocodiles present on the farm were free from infection was determined using the total number of farm born individuals tested (1,618) and all individuals tested (1,972) from 2008 2013 as the sample sizes. Analysis was performed using Freecalc (Cameron et al., 1998) with assumptions of 70% and 100% for the sensitivity and specificity of the artificial digestion method employed (Gamble et al., 2000), respectively, with a 95% confidence limit. The design prevalence for proof of freedom was set at 0.3%. 3. Results The results from testing crocodiles for infection with Trichinella are presented in Table 1. A total of 3,268 crocodiles were tested between the last few months of 2002 and the first seven months of 2013. They consisted of 1,065 wild born (518 males and 547 females) and 2,203 farm born animals (1,267 70 males and 931 females, plus 5 with no indication of sex). 71 72 The first crocodile found to be infected with T. papuae (5.0 larvae per g) was in the second farm cull of 2003; it arrived at the farm from Kikori in 2000. Between 2003 and 2007, 167 crocodiles originating from Page 3 of 11

4 73 74 Kikori were tested on reaching a size suitable for slaughter, 83 (50%) being positive for Trichinella. More male than female crocodiles (98:69) were tested, with 54.1% of males and 43.5% of females being 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 positive. In 2005 a few crocodiles (numbers in brackets) originating from Ambunti (4), Angoram (1), Murik Lakes (1), Wewak (4), Popondetta (1) and Kimbe (1) tested positive, as well as some (5) born on the farm. A further few were positive in 2006 and 2007; Ambunti (1), Timbunke (1), Popondetta (2) and farm born (2) (Fig. 1). No crocodile tested positive for Trichinella after 2007. There was a significantly greater proportion of wild born male crocodiles infected with Trichinella compared to wild born female crocodiles 9.5% (95% CIs: 7.3 12) and 5.2% (95% CIs: 3.7 7.1) respectively (p<0.05). The probability of observing no infected crocodiles in a sample of 1,618 farm born crocodiles and 1,972 farm and wild born crocodiles (total slaughtered from 2008 onwards), from a population with a prevalence of Trichinella infection of 0.3%, was 0.033 and 0.00 respectively. 4. Discussion and conclusion The discovery of infection in a few crocodiles born on the farm, as well as in crocodiles from localities other than Kikori, presented a potential problem for the farm because it suggested that infection was occurring within the farm. All crocodiles at the farm, both purchased and farm born, are fed exclusively on fresh poultry offal from the neighbouring chicken processing plant Infection was not acquired from this food because T. papuae is not infective to chickens (La Rosa et al., 2001). Feeding of animals with recycled raw crocodile meat has never been practiced on the farm. 93 94 95 As animals take 3 to 4 years at the farm to attain a suitable commercial size, and the farm stopped purchasing crocodiles from Kikori in 2003, the last crocodile on the farm from Kikori was killed in 2007. It was also the year when the last crocodile from any source was found to be positive for T. papuae. This Page 4 of 11

5 96 97 means that the last non Kikori crocodile tested positive four years after the last animal was purchased from Kikori. It also means that the first non Kikori animals to test positive in 2005 would have been 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 infected in 2001 or 2002, before testing had started. It was the practice at the farm up to sometime in 2004 for all juvenile crocodiles, wherever they came from, to be placed together in the same pens and for them to remain mixed for two or more years. Although individually tagged, neither their numbers nor health status were monitored regularly. Young crocodiles, therefore, continued to be mixed during 2003, when purchase of Kikori crocodiles ceased, and for part of 2004. It is significant that no crocodile, although in mixed company as a juvenile during 2004, was found infected with Trichinella when killed 3 or 4 years later. With the propensity of even young crocodiles for aggression, it is common knowledge that weak or sick individuals can be attacked and eaten; Huchzermeyer (2003) states cannibalism is a normal occurrence in crocodiles. The observation that a higher proportion of male crocodiles was infected with Trichinella compared to females is not unexpected, if juvenile males have the same tendency as adults to be more aggressive compared to females. The information available suggests strongly that infection within the farm was spread through cannibalism through eating the flesh of juveniles that died or by killing and eating live animals as a result of the now discontinued practice of housing farm born and village acquired crocodiles together. The small numbers of infected wild and farm born non Kikori, crocodiles, and the scattered original locations of the infected wild born animals, would appear to support this explanation. Furthermore, the 115 116 117 period of infection of these non Kikori animals can be linked to the time when young Kikori crocodiles were brought into the farm and the subsequent lack of infection in later years following cessation of Kikori purchases. Cannibalism probably took place soon after arrival at the farm, as newly purchased Page 5 of 11

6 118 119 animals sometimes fail to adapt to the new environment, becoming weak and vulnerable to attack by stronger individuals. 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 Purchase of young crocodiles from Kikori had been taking place for many years prior to the discovery of infection in 2003. But finding no infection amongst non Kikori crocodiles during 2003 and 2004, suggests that, although some Kikori crocodiles bought in 1999 or 2000 were infected as shown by positive results in those killed in 2003 and 2004 there is no evidence that cannibalism of infected animals occurred on the farm in those years. There is no means of ascertaining if infection was present in earlier years. The failure to detect any infected animals in the farm and wild born crocodiles tested from 2008 onwards is sufficient evidence to prove that the farm population is free from Trichinella infection at a design prevalence of 0.3%. This observation has been confirmed statistically using Freecalc, which calculates the exact probability of detecting infected animals in a population given the sensitivity and specificity of the tests used, the population size and design prevalence (lowest level detectable). The design prevalence in this case was set at a level below which transmission would not be sustained. This can be attributed to crocodiles no longer being purchased from Kikori since 2003 and the cessation since 2004 of the practice of placing all young animals together, irrespective of their place of origin. Conflict of interest statement Two of the authors, E.L. and W.S., are employed by the crocodile farm. 137 138 139 Acknowledgements The cooperation of the staff and management of Mainland Holdings crocodile farm is greatly appreciated. We thank the Australian Centre for International Agricultural Research (ACIAR) for Page 6 of 11

7 140 141 providing the funding for laboratory equipment and reagents as part of project AH/2001/054. Our thanks go also to Aaron Uforty, NAQIA, for assistance in preparing the map. 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 References Cameron, A.R., Baldock, F.C., 1998. A new probability formula for surveys to substantiate freedom from disease. Prev. Vet. Med. 34, 1 17. Gamble, H.R., Bessonov, A.S., Cuperlovic, K., Gajadhar, A.A., Knapen, F. van, Noeckler, K., Schenone, H., Zhu, X., 2000. International Commission on Trichinellosis: Recommendations on methods for the control of Trichinella in domestic and wild animals intended for human consumption. Vet. Parasitol. 93, 393 408. Huchzermeyer, F.W., 2003. Crocodiles: Biology, Husbandry and Disease. CABI Publishing, Wallingford, UK, 337pp. Khumjui, C., Choomkasien, P., Dakumyoy, P., Kusolsuk, T., Kongkaew, W., Chalamaat, M., Jones, J.L., 2008. Outbreak of trichinellosis caused by Trichinella papuae, Thailand, 2006. Emerg. Inf. Dis. 14, 1913 1915. Kusolsuk, T.,Kamonrattanakun, S., Wesanonthawech, A., Dekumyoy, P.,Thaenkham, U., Yoonuan, T., Nuamtanong, S., Sanguankiat, S., Pubampen, S., Maipanich, W., Panitchakit, J., Marucci, G., Pozio, E., Waikagul, J., 2010. The second outbreak of trichinellosis caused by Trichinella papuae in Thailand. Trans. R. Soc. Trop. Med. Hyg. 104, 433 437. 158 159 160 La Rosa, G., Marucci, P., Zarlenga, D.S., Pozio, E., 2001. Trichinella pseudospiralis populations of the Palearctic region and their relationship with populations of the Nearctic and Australian regions. Int. J. Parasitol. 31, 297 305. Page 7 of 11

8 161 162 Mukaratirwa, S., Foggin, C.M., 1999. Infectivity of Trichinella sp. isolated from Crocodylus niloticus to the indigenous Zimbabwean pig (Mukota). Int. J. Parasitol. 29, 1129 1131. 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 Owen, I.L., Gomez Morales, M.A., Pezzotti, P., Pozio, E., 2005. Trichinella infection in a hunting population of Papua New Guinea suggests an ancient relationship between Trichinella and human beings. Trans. R. Soc. Trop. Med. Hyg. 99, 618 624. Pozio, E., Foggin, C.M., Marucci, G., La Rosa, G., Sacchi, L., Corona, S., Rossi, P., Mukaratirwa, S., 2002. Trichinella zimbabwensis n.sp. (Nematoda), a new non encapsulated species from crocodiles (Crocodylus niloticus) in Zimbabwe also infecting mammals. Int. J. Parasitol. 32, 1787 1799. Pozio, E., Owen, I.L., La Rosa, G., Sacchi, L., Rossi, P., Corona, S., 1999. Trichinella papuae n. sp. (Nematoda), a new non encapsulated species from domestic and sylvatic swine of Papua New Guinea. Int. J. Parasitol. 29, 1825 1839. Pozio, E., Owen, I.L., Marucci, G., La Rosa, G., 2004. Trichinella papuae in saltwater crocodiles (Crocodylus porosus) of Papua New Guinea. Emerg. Inf. Dis. 10, 1507 1509. Pozio, E., Owen, I.L., Marucci, G., La Rosa, G., 2005. Inappropriate feeding practice favors the transmission of Trichinella papuae from wild pigs to saltwater crocodiles in Papua New Guinea. Vet. Parasitol. 127, 245 251. Webster, P., Malakauskas A., Kapel, C.M., 2002. Infectivity of Trichinella papuae for experimentally infected red foxes (Vulpes vulpes). Vet. Parasitol. 105, 215 218. Page 8 of 11

9 179 180 Fig. 1. Map of Papua New Guinea indicating coastal provinces and locations mentioned in the text, and including the number of crocodiles that tested positive for Trichinella papuae, together with the 181 182 183 number of crocodiles tested from each locality or source, during the period 2002 to 2013. Note: indicates location of crocodile farm. Page 9 of 11

10 183 184 185 186 Table 1. The proportion (%) of wild born and farm born crocodiles from Mainland Holdings Crocodile Farm, PNG, that were infected with Trichinella papuae from 2002 2013 (95% confidence intervals in parentheses). Wild born animals Farm born animals Year No. test No. pos % pos (95% CIs) No. Test No. pos % pos (95% CIs) 2003 69 16 23.2 (13.9, 34.9) 30* 0 0 2004 116 37 31.2 (23.6, 41.2) 31 0 0 2005 222 39 17.6 (12.8, 23.2) 109 5 4.6 (1.5, 10.4) 2006 196 6 3.1 (1.1, 6.5) 149 1 0.7 (0, 2.0) 2007 108 1 0.9 (0, 5.1) 266 1 0.4 (0, 2.1) 2008 86 0 0 364 0 0 2009 106 0 0 397 0 0 2010 101 0 0 303 0 0 2011 23 0 0 202 0 0 2012 8 0 0 262 0 0 2013 30 0 0 90 0 0 TOTAL 1065 99 9.3 (7.6, 11.2) 2203 7 0.3 (0.1, 0.7) *9 of these were tested in late 2002. Page 10 of 11

Accepted Manuscrip Page 11 of 11 Figure