Health and disease in Red-crowned Parakeets (Cyanoramphus novaezelandiae) on Tiritiri Matangi Island; causes of feather loss and implications for conservation managers
Dr Bethany Jackson BVSc MVS (Con Med) Murdoch University
A dissertation submitted to Murdoch University in fulfillment of the requirements of a Doctor of Philosophy
Supervisors Assoc Prof Kris Warren BSc, BVMS (Hons), PhD, Dip ECZM (Wildlife Population Health) Dr Richard Jakob-Hoff BSc (Hons), BVMS, MANZCVS (Wildlife Medicine) Dr Arvind Varsani BSc, PhD
Dr Carly Holyoake BSc BVMS PhD Prof Ian Robertson BVSc, PhD, MACVSc (Epidemiology)
November 2014
“This report, by its very length, defends itself against the risk of being read” Winston Churchill
ii
I declare that this thesis is my own account of my research and contains as its main content, work which has not been previously submitted for a degree at any tertiary education institution.
Dr Bethany Jackson BVSc MVS (Con Med) December 2014
iii ACKNOWLEDGEMENTS
Like most achievements in life, this project has come about through the generosity and support of a great many people, professionally and personally.
Thank you!!! To Kris Warren, Richard Jakob-Hoff, Arvind Varsani, Carly Holyoake, and
Ian Robertson for inspiration, guidance, support, feedback, encouragement, and a collective sense of humour when needed.
To the Auckland Zoo and the Auckland Zoo Conservation Fund for their financial, logistical and professional support of this project, as well as me, through the residency program. It was a pleasure to work for an organisation full of passionate staff and with a clear vision of the need to connect zoos, and visitors, with in situ conservation projects and activities. An enormous thank you must go to the very many Auckland
Zoo keepers and staff who provided such capable field assistance during the work, as well as some sensational cooking and entertainment packages. Also to the zoo maintenance staff for the plastic nest box extravaganza.
To the Supporters of Tiritiri Matangi Island, an incredible and enthusiastic bunch of genuinely lovely people with a passion for conservation of New Zealand species. They provided support in the field, as well as feedback into the design and implementation of the study based on their vast current and historical understanding of the island.
Volunteer groups like the Supporters of Tiritiri Matangi Island are fantastic ambassadors for what is achievable in conservation if you have a public that cares, which fortunately New Zealand does.
To the Department of Conservation rangers and Warkworth area office staff, for providing substantial support for field logistics on Tiritiri Matangi Island and Hauturu-
iv o-Toi/Little Barrier Island, including in kind funding. The project would not have been possible, or as enjoyable, without the assistance from the talented and dedicated DOC rangers and field staff on both islands, especially Dave Jenkins, Daryl Stephens, Jess
Clark, Jason Campbell, Nichy Brown and the effervescent Richard, Leigh, Liam and
Mahina Walle.
To Allen Heath, one of the few remaining experts in morphological identification of parasites in New Zealand, and a really wonderful human being. Thank you for being as
(if not more?) excited about the mites, which quite frankly turned me into a rather dull dinner party guest (Mary VA will testify). I feel incredibly fortunate to have had the opportunity to learn from and work with you.
Thank you also to the following organisations and individuals for funding and support, and believing in New Zealand species and their conservation:
360 Discovery Ferries, Bivouac Outdoor, Marley Plastics, Forest and Bird New Zealand,
New Zealand Veterinary Pathology, Massey University especially Brett Gartrell, Laryssa
Howe and Luis Ortiz-Catedral, Long Bay College Students (who so cheerfully made the wooden nest boxes!).
There are a few creatures that need to be thanked individually, for friendship and logistical support during the process, recording sighs of despair and lives saved, and enduring the endless talk about mites. They are Natalie Ann Sullivan, Danae Moore,
Kate McInnes, Emma Wells, Clio Reid, Mary Van Andel, Stephanie Jervis, Kirsten Derry,
Danny White, Peter Fraser, Mikaylie Wilson, Simon Fordham, Morag Fordham, John
Stewart, Kay Milton, and Mary Ann Rowland.
v ABSTRACT
Research into the diseases of free–ranging wildlife requires targeted surveillance for pathogen(s) of interest, however also relies on reference data of general health indicators against which findings can be interpreted. Wildlife reintroductions and translocations in New Zealand introduce specific disease risks related to spread of disease agents from the source site, exposure to novel disease at the destination, or changes in transmission factors for existing diseases at the destination. The discovery of Beak and feather disease virus (BFDV) causing clinical disease in wild Red-crowned
Parakeets (RCP, Cyanoramphus novaezelandiae) intended for translocation in 2008 led to an increased interest from conservation managers in the pathogens affecting this species.
This study aimed to investigate health and disease in a free-ranging population of RCP experiencing feather loss, with an epidemiological focus to determine temporal trends in disease prevalence, and infer risk factors for disease expression. We captured 229 individuals over 5 sampling sessions between 2011-2013, including 4 surveys on Tiritiri
Matangi Island (n=184), and one survey on Hauturu-o-Toi/Little Barrier Island (LBI)
(n=45).
Normal haematological and biochemical reference ranges were described. The relatively high creatine kinase results highlighted a potential susceptibility to capture myopathy in this species, and warrant further investigation. Comparison of DNA- based sexing results to beak measurements demonstrated high concordance between the two sexing methods.
vi We investigated BFDV in the population by PCR of blood and feather, and found a low total prevalence on Tiritiri Matangi Island of 1.09% (0.1-3.9%) for 2011-12, and on
Hauturu-o-Toi/LBI of 4.4% (0.5%-15.1%) in 2013. Screening by PCR of opportunistic samples from the Wellington region of the North Island, and phylogenetic analysis of the full viral genome sequences from all positive samples, revealed ongoing evidence of viral flow between RCP and Eastern Rosellas (Platycercus eximius) in the Hauraki
Gulf/Auckland region, with separate but closely related strains from the Wellington region. These findings, combined with the first report of seroprevalence data for a
New Zealand parrot using the haemagglutination inhibition test, suggest RCP may be a dead-end or spill over host for BFDV, with implications for a downgrading of the conservation threat this pathogen currently poses.
The collection of skin biopsies in all birds during the 2012 and 2013 (n=135) following an outbreak of feather loss led to the discovery of a knemidokoptinid mite associated with mange in this species. The mite, Procnemidocoptes janssensi has only been previously described once from a lovebird (Agapornis nigrigenis) in Zambia in 1967.
We present the first report of the mite in New Zealand, including histopathological stages of infestation, and the first epidemiological study on mange in a wild parrot globally. Relative mite abundance (number of mites per skin biopsy) was found to be associated with likelihood of feather loss and clinical signs. Findings were also consistent with the presence of a carrier or endemic state, with emergence of clinical disease and epizootics triggered by as yet unknown host, environment or mite factors.
Finally, we conducted a nesting study using natural and artificial nests on Tiritiri
Matangi Island from 2012-2013, to infer disease risks for nestling and fledgling RCP, and describe baseline health parameters for this group. Overall nesting success was
vii low, and likely to be related to the drought experienced that year. Nest mites were detected in 64% (95%CI: 41-83%) of nests, however their presence did not significantly affect fledging success or numbers. Chicks from nest mite positive nests had higher mean absolute heterophil counts (p=0.04), suggesting an inflammatory response to the presence of these mites, and a potential fitness cost that warrants further study. We did not detect BFDV in any nests studied, and therefore cannot infer the impact of this pathogen on this age class, although the study provided further evidence BFDV is present at a low prevalence on Tiritiri Matangi Island.
viii Publications from or related to this thesis
Peer reviewed:
Jackson B, Harvey C, Galbraith J, Robertson M, Warren K, Holyoake C, Julian L, and
Varsani A. 2014. Clinical beak and feather disease virus infection in wild juvenile eastern rosellas of New Zealand; biosecurity implications for wildlife care facilities,
New Zealand Veterinary Journal, 62(5): 297-301.
Jackson B, Lorenzo A, Theuerkauf J, Barnaud A, Duval T, Guichard P, Bloc H, Baouma A,
Stainton D, Kraberger S, Murphy S, Clark N, Dillon C, Knight T, and Varsani A. 2014.
Preliminary surveillance for beak and feather disease virus in wild parrots of New
Caledonia: implications of a reservoir species for Ouvea Parakeets. Emu 114(3): 283-
289.
Massaro M, Ortiz-Catedral L, Kurenbach B, Kearvell J, Kemp J, Van Hal J, Elkington E,
Galbraith J, Taylor G, McInnes K, Heber S, Steeves T, Walters M, Shaw S, Potter J,
Farrant M, Brunton DH, Hauber M, Jackson B, Moorhouse R and Varsani A. 2012.
Molecular characterisation of Beak and feather disease virus (BFDV) in New Zealand and its implications for managing an emerging, infectious disease. Archives of Virology
157(9): 1651-63.
Julian L, Lorenzo A, Chenuet JP, Borzon M, Marchal C, Vignon L, Collings DA, Walters
M, Jackson B, and Varsani A. 2012. Evidence of multiple introductions of beak and feather disease virus into the Pacific islands of Nouvelle-Caledonie (New Caledonia).
Journal of General Virology 93(11): 2466-72.
ix Jackson B, Heath A, Harvey C, Holyoake C, Jakob-Hoff R, Varsani A, Robertson I, and
Warren K. 2014. First report of knemidokoptinid (Epidermoptidae: Knemidokoptinae) mite infestation in wild red-crowned parakeets (Cyanoramphus novaezelandiae); correlations between macroscopic and microscopic findings. Journal of Wildlife
Diseases. In press.
Jackson B, Varsani A, Holyoake C, Jakob-Hoff R, Robertson I, McInnes K, Empsom R,
Gray R, Nakagawa K, and Warren K. 2014. Emerging infectious disease or evidence of endemicity? A multi season study of Beak and feather disease virus in wild Red- crowned Parakeets (Cyanoramphus novaezelandiae). In Review: Archives of Virology.
The following manuscripts have been prepared for submission:
Jackson B, Jakob-Hoff R, Holyoake C, Robertson I, Varsani A, and Warren K. 2014.
Haematology and biochemistry reference ranges for wild Red-crowned Parakeets
(Cyanoramphus novaezelandiae) of New Zealand; creatine kinase results suggest risks of capture myopathy. In review New Zealand Veterinary Journal.
Jackson B, Heath A, Holyoake C, Jakob-Hoff R, Varsani A, Robertson I, and Warren K.
2014. Impact of haematophagous nest mites (Dermanyssus sp.) and knemidokoptinid mange mites on reproductive parameters and fledgling haematology in wild Red- crowned Parakeets (Cyanoramphus novaezelandiae).
Jackson B, Heath A, Harvey C, Holyoake C, Jakob-Hoff R, Varsani A, Robertson I, and
Warren K. 2014. Epizootic mange in wild Red-crowned Parakeets; seasonal trends and risk factor analysis.
x Media and non peer-reviewed publications
Jackson, B. Science and the protection of our native wildlife. Bivouac mail. 2011.
Jackson, B. “That’s Science” interview. B-FM. 4 August 2011.
Penman, M. Students get behind project. North Shore Times. 2 December 2011.
Jackson, B. 2012. Can you spot the difference? Causes of feather loss in kakariki on
Tiritiri Matangi. Dawn Chorus. 88.
Jackson, B. Researchers studying spread of disease affecting kakariki. Interview with
W. Hines. Radio National. 30 March 2012.
Jackson, B. and Raisin, C. 2012. A tale of two psittacines. PsittaScene. 24(3).
Jackson, B. Kakariki struggle through summer. Interview with A. Bourn. 3-News
National. 18 February 2013.
Jackson, B. Fight to save New Zealand’s parakeets. Interview with W. Hines. One News national bulletin. 15 June 2013.
Jackson, B. 2013. What caused our kakariki to start going bald? Dawn Chorus. 94.
xi Conferences and presentations
Jackson, B. Wildlife rehabilitation; Disease surveillance at the frontline. Wildlife
Rehabilitators Network of New Zealand. Auckland, New Zealand. 14-15 July 2012.
Jackson, B. Naked Kakariki! What is causing feather loss in these vulnerable New
Zealand parakeets? Auckland Zoological Society Seminar Series. Auckland Zoo, New
Zealand. 25 July 2012.
Jackson, B. Disease surveillance in New Zealand: the role of sanctuaries. Sanctuaries of
New Zealand workshop. Inglewood, New Zealand. July 2012
Jackson, B., Heath, A., Holyoake, C., Jakob-Hoff, R., Robertson, I., Varsani, A. and
Warren, K. A mite-y problem: What is causing feather loss in Red crowned Parakeets on Tiritiri Matangi Island? Wildlife Disease Association – Australasia Conference.
Grampians, Australia, 29 September – 4 October 2013.
Jackson, B., Heath, A., Harvey, C., Warren, K., Holyoake, C., Jakob-Hoff, R., Robertson,
I., and Varsani, A. Epizootic mange in kakariki; biosecurity indiscretion or assisted self- introduction? New Zealand Society for Parasitology Conference and Annual Meeting.
Palmerston North, New Zealand, 20 - 22 October 2013.
Jackson, B., Heath, A., Varsani, A., Warren, K., Jakob-Hoff, R., Holyoake, C. and
Robertson, I. Unmasking the culprit - a veterinary approach to conservation threats in wild kakariki. Australasian Wildlife Management Society Conference. Palmerston
North, New Zealand, 20 - 22 November 2013.
xii Glossary
Pathogen Infectious agents capable of causing disease, including viruses, macroparasites, bacteria, fungi, protozoans and prions
Iwi Māori social unit, usually referring to a tribe
Hapū Māori social unit, referring to a group of families, multiple hapū together form iwi.
Treaty of Waitangi The treaty signed in 1840 between Britain and Māori chiefs of the North Island, giving Britain sovereignty over New Zealand
Taonga Treasured by Māori
Abbreviations
BA Bile acids
BFDV Beak and feather disease virus
CI Confidence Interval
CK Creatine kinase
DNA deoxyribonucleic acid
DOC Department of Conservation (New Zealand)
DRA Disease Risk Analysis
EWCC Estimated white cell count
HE Haematoxylin and Eosin
HR Heart rate
IU International units km kilometre
LBI Little Barrier Island
xiii MPI Ministry of Primary Industries (New Zealand)
OFP Malherbe’s or Orange-fronted Parakeet, Cyanoramphus malherbi
OR Odds ratio p p-value or probability value
PCR Polymerase chain reaction
PCV Packed cell volume
RCP Red-crowned Parakeet, kakariki, Cyanoramphus novaezelandiae
RR Respiratory rate sd standard deviation
SoTM Supporters of Tiritiri Matangi Island (community group)
TP Total Protein
TTM Tiritiri Matangi Island
UA Uric acid
WBC White Blood Cell
YCP Yellow-crowned Parakeet, Cyanoramphus auriceps
N.B. Virus nomenclature
Species names for viruses are in the form and typographical style nominated by “The International Code of Virus Classification and Nomenclature February 2013” (http://www.ictvonline.org/codeOfVirusClassification.asp, accessed 27 November 2014).
Accordingly, species names are italicised, with the first word in the name capitalised, and subsequent words lower case unless proper nouns (e.g. Beak and feather disease virus).
xiv
N.B. Common names for bird species
Māori common bird names are not capitalised (e.g. kākā), however English common names are spelt according to the International Ornithologist’s Union world bird list version 4.4, which uses capitalisation of common names (e.g. Red-crowned Parakeet) (http://www.worldbirdnames.org, accessed 27 November 2014).
xv Table of Contents
Declaration iii Acknowledgements iv Abstract vi Publications, media and presentations ix Glossary and abbreviations xiii Table of contents xvi List of figures xix List of tables xxiii Chapter 1: Introduction 1 1.1. Background to study 2 1.2. Contemporary conservation in New Zealand 3
1.2.1. History of species declines and current solutions 3
1.2.2. Disease risks associated with animal movements 5
1.2.3. Health and disease screening 7 1.3. Red crowned Parakeets 9
1.3.1. Distribution and threat status 9
1.3.2. Biology 9
1.3.3. Tiritiri Matangi Island and Hauturu-o-Toi/LBI 10 1.4. Diseases of wild Red-crowned Parakeets 13
1.4.1. Infectious diseases and parasites detected in New Zealand 13
1.4.2. Disease surveillance on Tiritiri Matangi Island and Hauturu-o-Toi/LBI 15 1.5. Disease implications of trade for wild RCP 16
1.5.1. Trade and the establishment of wild exotic parrots in New Zealand 16
1.5.2. Diseases of domestic parrots in New Zealand 18
1.5.3. Disease implications of the parrot trade in New Zealand 19
1.5.4. Disease risk management for conservation 21
xvi 1.6. Aims of this study 21 Chapter 2: General Methods 23 2.1. Mist netting 24 2.2. Location of captures 24 2.3. Handling and removal 26 2.4. Bycatch 26 2.5. Anaesthesia and sample collection 27 2.5.1. Set up of field anaesthetic station 27 2.5.2. Roles of personnel 28 2.5.3. Anaesthetic protocol 29 2.5.4. Sample collection 30 2.6. Physical exam, morphometric measurements and photos 32 2.7. Disinfection/biosecurity 34 Chapter 3: Haematology, biochemistry and physical reference ranges 36 3.1. Introduction 37 3.2. Methods 39 3.3. Results 43 3.4. Discussion 48 3.5. Conclusions 54 Chapter 4: Beak and feather disease virus in Red-crowned Parakeets 55 4.1. Introduction 56 4.2. Methods 61 4.3. Results 65 4.4. Discussion 71 4.5. Conclusions 77
xvii Chapter 5: Histopathology of mange in wild Red-crowned Parakeets 79 5.1. Introduction 80 5.2. Methods 83 5.3. Results 87 5.4. Discussion 96 5.5. Conclusions 103 Chapter 6: Epidemiology of mange in wild Red-crowned Parakeets 105 6.1. Introduction 106 6.2. Methods 108 6.3. Results 113 6.4. Discussion 123 6.5. Conclusions 133 Chapter 7: Disease risks for nesting success in Red-crowned Parakeets 134 7.1. Introduction 135 7.2. Methods 138 7.3. Results 147 7.4. Discussion 155 7.5. Conclusions 165 Chapter 8: General Conclusions 167 8.1. Health and disease in conservation management 168 8.2. Updated knowledge of disease in wild Red-crowned Parakeets 171 8.3. Methods for wildlife disease surveillance 174 8.4. Biosecurity and public health for wildlife research 177 8.5. Research recommendations 180
References 184 Appendices 220
xviii List of figures
Figure 1.1 Map of New Zealand with outlined area of North Island enlarged in
insert, showing the two study sites; Tiritiri Matangi Island and Hauturu-
o-Toi/Little Barrier Island.
Figure 2.1 Mist net 9m x 3m erected on 5m poles and placed in flight path at
Emergency Landing, Tiritiri Matangi Island.
Figure 2.2 Map showing sites of captures and sampling of Red-crowned Parakeets
on Tiritiri Matangi Island.
Figure 2.3 Map showing site of captures and sampling of Red-crowned Parakeets
on Hauturu-o-Toi/LBI.
Figure 2.4 Sampling a Red-crowned Parakeet on Tiritiri Matangi Island. In the
foreground a Red-crowned Parakeet can be seen under anaethesia,
with a nurse monitoring vital signs and the author to the left of the
screen. The end of the table in the foreground is considered dirty and
thoroughly cleaned between sampling birds, the far end of the table is
considered clean with samples and data management being conducted
by Auckland Zoo keeping staff.
Figure 2.5 Examples of standardised photos obtained of Red-crowned Parakeets
during the study.
Figure 3.1 Correlation between creatine kinase levels (U/l) and time from capture
to sampling (min) in wild Red-crowned Parakeets on Tiritiri Matangi
Island, New Zealand.
xix Figure 3.2 Comparison of beak measurements (culmen width and length in mm)
with PCR-based DNA sexing results for Red-crowned Parakeets on Tiritiri
Matangi Island and Hauturu-o-Toi/LBI. Individuals sexed as males by
PCR are indicated by white dots, females by grey dots.
Figure 4.1 Map of the North Island of New Zealand showing sites where samples
for Beak and feather disease virus screening were obtained from Red-
crowned Parakeets.
Figure 4.2 Maximum likelihood phylogenetic tree of the full genomes of BFDV
isolates on the North Island of New Zealand (BFDV-A strain), showing
the relationship between Eastern Rosella (purple line) and Red-crowned
Parakeet (green line) isolates. Pairwise identities (%) are presented on
the right of the tree, with all isolates sharing ≥95% pairwise identity.
BFDV isolates from this study are in bold font and year of sampling is
provided in brackets after the GenBank accession numbers.
Figure 5.1 Stages of infestation with Procnemidocoptes janssensi in Red-crowned
Parakeet skin, including normal skin and a whole mite specimen. 5.1a)
Normal skin demonstrating the thin epidermis, basket-weave keratin
layer, and minimal cellularity to dermis. HE. 5.1b) Widening of feather
follicle due to presence of adult female mite at base (white arrow), with
associated moderate acanthosis and hyperkeratosis. Faeces evident
(black arrow) proximal to the mite. Gnathosoma (A) visible with mouth
parts. HE. 5.1c) One large adult female mite (black arrow) with
chitinous (A) and developing (B) larval forms inside. Body shape and
xx short legs of adult female support identification as knemidokoptinid
mite. Larvae are also evident adjacent to the female mite (C). Mites are
overlying the dermal papilla (white arrow) of the feather follicle, with
moderate surrounding acanthosis and hyperkeratosis. HE. 5.1d)
Marked mixed mononuclear and granulocytic diffuse dermal
inflammation (white arrow) subjacent to mite penetrating the
epidermis (black arrow). Marked acanthosis evident in adjacent
epidermis. HE. 5.1e) Pretarsal stalk and pulvillus (arrow) evident on
second leg of adult female mite, definitive for P. janssensi compared
with other knemidokoptinids. Female mite has developing embryo (A).
HE. 5.1f) Cleared whole adult female mite with identifying features of
P. janssensi, including chitinous larval form inside (A). Hoyer’s medium.
Figure 5.2 Red-crowned Parakeets with feather loss on Tiritiri Matangi Island.
Female with grade 3 feather loss affecting >50% of head, and mild scale
formation (5.2a), and male with grade 3 feather loss, ventral neck and
keel severely affected (5.2b).
Figure 5.3 Density histogram for mite counts from skin biopsies of red-crowned
parakeet on Tiritiri Matangi Island and Hauturu-o-Toi/LBI 2012-2013.
Figure 7.1 Map of Tiritiri Matangi Island showing location of nest boxes and
successful nests, including plastic boxes (), wooden boxes (), and
natural nests (). Successful nests (chicks fledged) are filled in with
black, with number of chicks fledged indicated by the number to the top
right of the nest.
xxi Figure 7.2 Wooden nest boxes made of macrocarpa for Red-crowned Parakeets on
Tiritiri Matangi Island.
Figure 7.3 Components of plastic nest box design for Red-crowned Parakeets on
Tiritiri Matangi Island. 7.3a – Hinged door with inlaid rim for
waterproofing, 7.3b – removable wooden ladders leading to exit, with
ventilation holes visible at the top of the nest box.
Figure 7.4 Plastic nest box for Red-crowned Parakeets in situ on Tiritiri Matangi
Island.
Figure 7.5 Inside plastic nest box with wooden ladders visible and 8 eggs on
wooden shavings used as nesting substrate.
xxii List of tables
Table 3.1 Selected haematology and biochemistry reference ranges for Red-
crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-Toi/Little
Barrier Island.
Table 3.2 Descriptive statistics by sex for standard morphometric measurements
(mm) from normal Red-crowned Parakeets on Tiritiri Matangi Island and
Hauturu-o-Toi/Little Barrier Island
Table 3.3 Descriptive statistics by sex and location for weights (gm) of normal
Red-crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-
Toi/Little Barrier Island
Table 4.1 PCR and Haemagglutination Inhibition prevalence results for BFDV
testing of Red-crowned Parakeets at various locations on the North
Island of New Zealand during the study period. PCR positive results are
from blood and/or feather samples.
Table 4.2 Individual Haemagglutination Inhibition (HI) titre results from testing of
Red-crowned Parakeets on Hauturu-o-Toi/Little Barrier Island and
Tiritiri Matangi Island
Table 4.3 BFDV positive Red-crowned Parakeets from Tiritiri Matangi Island and
Hauturu-o-Toi/Little Barrier Island, including sex, location, blood and
feather PCR results, and HI results.
Table 4.4 Selected reported BFDV prevalence data for wild exotic and native
parrots in New Zealand
xxiii Table 5.1 Avian families commonly infested by the Knemidokoptinae, with species
of mites found and area of the body affected during clinical disease.
Translated and adapted to reflect current nomenclature from Fain and
Elsen (1967).
Table 5.2 Prevalence and range data for Procnemidocoptes janssensi mites in skin
biopsies, including locations of mites found. ‘Overall mites’ indicates
proportion of skin biopsies (n=135) with mites observed. Intrafollicular
and extrafollicular mites are expressed as a proportion of the mite-
positive biopsies (n=79).
Table 5.3 Odds ratios, 95%CI and Fisher’s exact p-value for relationships between
outcome of combined hyperkeratosis/acanthosis and feather loss,
mites, or feather loss and mites combined.
Table 5.4 Species, stage and total numbers of mites found in skin biopsies from
the back of the head of six birds during cross sectional study on Tiritiri
Matangi Island.
Table 5.5 Adult, nymph and larval forms of Procnemidocoptes janssensi recovered
from serial 3mm biopsy sections of skin from the body of a deceased
RCP with mange
Table 6.1 Logistic regression model showing the relationship between total mites
in biopsy and the odds of mange.
Table 6.2 Prevalence of mange (Y/N) in Red-crowned Parakeets on Tiritiri Matangi
Island and Hauturu-o-Toi/LBI across years and seasons.
xxiv Table 6.3 Prevalence of mites (Y/N) in Red-crowned Parakeets on Tiritiri Matangi
Island and Hauturu-o-Toi/LBI, across seasons and years.
Table 6.4 Results for mean mite counts in biopsies of Red-crowned Parakeets by
sex and location
Table 6.5 Model contents of Generalised Linear Model (GLM) for mange in Red-
crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-Toi/Little
Barrier Island.
Table 7.1 Clutch size, live young and fledge rate (± sd where available) for all nest
types of Red-crowned Parakeets on Tiritiri Matangi Island 2012-13,
compared with previous studies.
Table 7.2 Comparison of clutch size, live young and fledge rate between plastic
nest boxes and natural nest sites for Red-crowned Parakeets nesting on
Tiritiri Matangi Island 2012-13.
Table 7.3 Prevalence data for infestation of Red-crowned Parakeet nests with
nest mites, and ectoparasitic mange in parents observed on nests, on
Tiritiri Matangi Island 2012-13.
Table 7.4 Comparison of reproductive parameters between nest mite-positive
and -negative nests of Red-crowned Parakeets on Tiritiri Matangi Island
2012-13.
Table 7.5 Haematology of fledging Red-crowned Parakeet chicks from nest mite-
positive and -negative nests on Tiritiri Matangi Island 2012-13.
xxv
CHAPTER 1 GENERAL INTRODUCTION
1 1.1 Background to study
In 2008, Beak and feather disease virus (BFDV) was detected for the first time in a wild native New Zealand parrot, during a translocation event of Red-crowned
Parakeets (RCP, Cyanoramphus novaezelandiae) (Ortiz-Catedral et al. 2009b). The affected birds were from Hauturu-o-Toi/Little Barrier Island (LBI), in the Hauraki Gulf of New Zealand (Figure 1.1). BFDV can cause significant morbidity or mortality in affected birds (Raidal 1995), and the virus is a well-recognised cause of infectious disease of parrots with conservation management implications for wild populations globally (Borthwick 2005). RCP on nearby Tiritiri Matangi Island had anecdotal reports of feather loss, however BFDV had not yet been detected on this island
(Massaro et al. 2012).
The implications of detecting BFDV in 2008 were significant for native New Zealand parrots for multiple reasons. New Zealand hosts many endemic and unique parrot species, including the critically endangered Malherbe’s Parakeet (Orange-fronted
Parakeet, OFP, Cyanoramphus malherbi) and kākāpō (Strigops habroptilus). Most species exist only in fragmented small populations on islands or in sanctuaries where exotic predators are managed or eradicated. Translocations and reintroductions form a necessary part of their conservation management. As a result of the detection of BFDV, a ban was instigated on native parrot movements to the wild (e.g. wild-wild transfers and captive-wild transfers) pending further information on the host and geographic spread of this virus in native parrots. A joint Ministry of Primary
Industry (MPI) and Department of Conservation (DOC) response was launched to investigate the threat posed by BFDV to native parrots (Pers. comm. Kate McInnes,
DOC). These steps highlighted the precautionary approach often adopted in wildlife
2 disease risk analyses due to a lack of baseline data and information on species- specific hazards or diseases of interest (Jakob-Hoff et al. 2014), as well as the need for further information on BFDV and other pathogens/parasites present in RCP in
New Zealand.
Figure 1.1: Map of New Zealand with outlined area of North Island enlarged in insert, showing the two study sites: Tiritiri Matangi Island and Hauturu-o-Toi/Little Barrier
Island.
1.2 Contemporary conservation in New Zealand
1.2.1 History of species decline and current solutions
The level of species endemism in New Zealand makes this country a unique and high priority conservation region at an international level (Myers et al. 2000), and imbues
3 a sense of urgency to native species protection. Whilst isolation allowed New
Zealand to develop a distinct and iconic faunal assemblage, it left these endemic species exposed and vulnerable to exotic invasions which were primarily associated with anthropogenic activities. New Zealand has suffered one of the worst extinction rates since humans arrived (ME 2007), as a result of the usual suite of threats; namely introduced mammalian predators, and habitat destruction and fragmentation (Saunders and Norton 2001). Consequently species are vulnerable to issues faced by small and isolated populations, including loss of genetic diversity and stochastic events, that may render them more susceptible to novel or existing diseases (de Castro and Bolker 2005).
New Zealand has obligations at many levels to mitigate biodiversity loss through conservation of its native fauna, including:
Being a signatory to the Convention on Biological Diversity, particularly “Article 8:
In Situ Conservation” (UNEP 1992).
Recognising the Treaty of Waitangi and a responsibility to protect biodiversity
interests of iwi and hapū, particularly taonga species (ME 2000).
Having a national responsibility as custodians of the land to restore and protect
the environment of New Zealand, including its native fauna, in accordance with
the Wildlife Act (1953) and the Environment Protection Authority Act (2011).
Preservation of ecosystem functions, goods and services that provide for
socioeconomic stability, and the sustainability of society (Nellemann and
Corcoran 2010).
4 The New Zealand Biodiversity Strategy (2000) was developed to address these obligations, following evidence from the State of New Zealand’s Environment report that loss of biodiversity was the most significant environmental issue for New
Zealand (ME 1997).
The native fauna of New Zealand are particularly vulnerable to introduced mammalian predators (including mustelids, rats, possums, cats, dogs and hedgehogs), having evolved in the absence of these or analogous species. In the mid
1980’s only 2,162 ha of New Zealand remained un-invaded by introduced predators
(Bellingham et al. 2009). This lead to the intensification of a conservation solution that focussed on the creation of introduced predator-free offshore and mainland island sites to safely house recovering native species (Saunders and Norton 2001,
Parker 2013). New Zealand is a world leader in island pest eradications and habitat restoration, with notable successes including the removal of rats from large areas, such as Campbell Island (11,257 ha) (Bellingham et al. 2009). A key component of this ‘ark’ solution is the movement of species between sites for genetic management, restoration of species assemblages, and insurance populations. Parker et al. (2006) reported more than 400 translocations of native species have occurred in New Zealand, the majority of these involving birds.
1.2.2 Disease risks associated with animal movements
Throughout history the potentially catastrophic effects of disease have been realised in both human and wildlife populations (Daszak and Cunningham 1999, Daszak et al.
2000, Lafferty and Gerber 2002). Significant examples include the Spanish flu pandemic of the early 20th century, with an estimated 50-100 million human deaths
5 worldwide (Johnson and Mueller 2002). There is evidence that an H1N1 influenza virus caused this flu outbreak, likely occurring as a ‘spill over’ event with avian origins (Dudley 2008). Just 20 years earlier the introduction of Rinderpest virus by infected cattle imported to South Africa, resulted in the death of an estimated 95% of exposed wild ungulates, with remodelling of the affected ecosystems both during and after the disease event (Lafferty et al. 2005). However, disease works at many levels, and sub-lethal effects may still cause population-scale changes (Spalding and
Forrester 1993). An animal with sub-lethal disease may lose normal predator avoidance behaviours, take greater risks, or may fail to reproduce. In this way, disease may be the proximate or ultimate cause of a decline, and may augment other threatening processes to lead species into extinction vortices.
Movement of animals between sites introduces disease risks from the source and destination sites (Sainsbury and Vaughan-Higgins 2012). These include introduction of novel pathogens or parasites from the individuals being moved to the recipient population, as well as the potential that new individuals will alter transmission dynamics of local or native pathogens (Cunningham 1996). There are also direct and indirect disease risks to sympatric species at the destination site, with potential consequences for ecosystems related to disease-mediated population changes
(Cunningham 1996, Sainsbury and Vaughan-Higgins 2012). Finally, the translocated individuals may be at risk of exposure to new pathogens at the destination site, which may put the success of the translocation at risk, incurring considerable time and financial costs. For these reasons, qualitative or quantitative disease risk analyses have been recommended to evaluate the likelihood and consequence of
6 disease spread during movements, as well as produce transparent methods for evaluating these risks and reaching consensus on whether to proceed (Jakob-Hoff et al. 2014).
Disease events associated with anthropogenic wildlife movements are increasingly recognised (Dobson and Lyles 2000, Schloegel et al. 2006, Skerratt et al. 2007) as a result of growing awareness of the risks and the more frequent inclusion of disease surveillance in wildlife research. Possibly the most significant current disease issue in relation to biodiversity conservation is the combined impacts of Batrachochytridium dendrobatidis (a fungus) and Ranavirus on global amphibian populations (Schloegel et al. 2010, Grogan et al. 2014). The legal and illegal trade in amphibians, captive breed for release programs, and translocations, have been identified as contributing to the spread of these pathogens in wild amphibians (Walker et al. 2008, Schloegel et al. 2010). Massive global declines and extirpations have been recorded in amphibians, and, significantly, chytrid fungus has now been listed as a notifiable disease by the World Animal Health organisation (OIE) (Schloegel et al. 2010).
1.2.3 Health and disease screening for translocations and reintroductions; considerations and New Zealand perspectives
For the reasons cited above, there is a clear need for a degree of health and disease screening of wildlife intended for translocation (including, at times, screening of individuals at the destination site) (Cunningham 1996). However methods for approaching these activities vary and are affected by factors that include; (i) cost:benefit assessments, (ii) baseline knowledge for the species, locations and systems of interest, (iii) availability and accuracy of testing for diseases in the target
7 species, and (iv) availability of expertise for obtaining and interpreting findings. An important element that may affect the initiation, direction and adoption of recommendations from any disease-related activity is risk perception, which is usually dictated by attitudes and priorities.
The translocation history in New Zealand, coupled with a general in-country awareness of the disease risks for wildlife translocations, has led to multiple efforts to address these risks and to establish baseline data for a range of species (Low et al.
2006, Parker et al. 2006, Ortiz-Catedral et al. 2009a, Baling et al. 2013). These efforts are reflected at multiple scales and across stakeholders, including the DOC permit process that often mandates disease screening; establishment of national disease surveillance and risk assessment capacity building workshops (Jakob-Hoff 2010); development of national databases of wildlife health and pathology data (DOC); and research undertaken through universities, Crown Research Institutes, local government and community groups. However, in spite of these efforts, available information is relatively species-specific, and often directed towards high profile or commonly translocated species. Further, disease screening is often limited to either known pathogens, or pathogens for which testing is available cost effectively in country. Recently the use of metagenomics has enabled the limited exploration of novel pathogens to determine the aetiology of enigmatic disease syndromes such as cloacitis in the kākāpō (White et al. 2014).
8 1.3 Red-crowned Parakeets
1.3.1 Distribution and threat status
Red-crowned Parakeets belong to the Cyanoramphus genus, comprised of 10 species and four sub-species (extant and extinct) spanning New Zealand and its off-shore islands, as well as New Caledonia and Norfolk Island (Boon et al. 2001). A review of the molecular systematics of the genus found Cyanoramphus species in New Zealand represent an ancient link with New Caledonia, and likely arrived from the latter region via Norfolk Island (Boon et al. 2001). Following the arrival of humans and associated threats including persecution as crop pests, RCP became extinct on the mainland and restricted to a few offshore islands (Greene 1998). They are listed as
“Vulnerable” by the IUCN (BirdLife International 2013), and “At Risk-Relict” by the
New Zealand Threat Classification System (Robertson et al. 2012). It should be noted that whilst a permit is required to hold this species within New Zealand, RCP are a popular domestic pet species globally.
1.3.2 Biology
RCP are a small, sexually dimorphic parakeet, generally weighing 70-100g with males larger than females (Forshaw 2010). Males normally have a wider and longer beak, thus beak measurements are often used to determine sex (Sagar 1988). Research into their foraging ecology (Greene 1998) and reproductive biology (Greene 2003,
Ortiz-Catedral and Brunton 2010a) has highlighted a greater flexibility in the use of resources for feeding and breeding than their mainland counterparts, the Yellow- crowned Parakeet (Cyanoramphus auriceps). Although described as a forest-dwelling
9 species (Greene 1998), RCPs appear to cope well in modified or regenerating habitats (Ortiz-Catedral and Brunton 2008, 2010b), and they nest in a wide variety of locations from the base of flax plants to seabird cliff nests, as well as traditional cavities in trees (Greene 2003, Ortiz-Catedral and Brunton 2010a). Whilst this flexibility serves them well in reintroductions to recovering sites that may lack old growth forest, their tendency to forage in open-spaces and to ground nest renders them more vulnerable to the effects of predation. Records suggest RCPs are more vulnerable to local extinction in the presence of browsers and predators, rather than predators alone (Greene 1998), highlighting the synergistic impact of reduced resources on this species. RCPs nest during the summer months, generally from
October to March, although annual fluctuations are reported (Ortiz-Catedral and
Brunton 2008), possibly in response to climate driven masting events of local food plants (Elliot et al. 1996). Clutches may be large (up to 9 eggs), although fledge rates are substantially lower, with reported ranges of 1.04-3.27 per nest (Ortiz-Catedral and Brunton 2008).
1.3.3 Tiritiri Matangi Island and Hauturu-o-Toi/Little Barrier Island
Location and history of RCP
Tiritiri Matangi (220 ha, S36˚ 36' 2", E174˚ 53' 24") is an open island sanctuary located in the Hauraki Gulf of the Auckland region (Figure 1.1). Habitat restoration from a large-scale replanting program between 1984-1994 returned young native forest to approximately 60% of the island (Galbraith and Cooper 2013). The island experiences a low annual rainfall and generally milder temperatures compared to the mainland, and is free of non-native predators following the eradication of the
10 kiore (Pacific rat, Rattus exulans) in 1993 (Rimmer 2009). Twelve avian species have been translocated to Tiritiri Matangi island, of which 11 have established breeding groups (Parker 2013). It remains a focus of research for multiple institutions and organisations including the Supporters of Tiritiri Matangi Island, a volunteer community group (Galbraith and Cooper 2013). The island also has a significant advocacy role for New Zealand conservation and native species with over 36,000 visitors annually (Rimmer 2009). Maori interests are represented by iwi including
Ngati Paoa, Ngati Wai, Ngati Manuhiri, Nga Manawhenua o Tamaki, Te Kawerau a
Maki, Te Marutuahu, Te Runanga o Ngati Whatua, Ngati Maru, Ngati Tamatera, Te
Patukirikiri, Ngai Tai Ki Tamaki, Ngati Te Ata, and Ngati Tamaoho.
RCP successfully established on Tiritiri Matangi Island following their introduction in the 1970’s from captive stock at Mount Bruce Wildlife Centre (McHalick 1998).
Transfers were under instruction from the Wildlife Service (now incorporated into
DOC), with three transfers in 1974 (n=35), 1975 (n=22) and 1976 (n=27) (McHalick
1998). The origins of the captive stock at Mount Bruce are less clear, and may have been from captive avicultural stock in the Auckland region (Pers. comm. C.R. Veitch, formerly Wildlife Service). Whether the original release was planned or otherwise has been debated in the literature (Parker 2013). Records from Mount Bruce clearly state the birds were intended for release on Cuvier Island, however were diverted to
Tiritiri Matangi Island following the failure of transport to Cuvier Island (Mount Bruce
Native Bird Reserve 1974), a fact supported by the wildlife officer who conducted the liberation (Pers. comm. C. Smuts-Kennedy). The origins of the captive birds (wild
11 stock or avicultural) have important implications for potential sources of any diseases found in the RCP population on Tiritiri Matangi Island.
Hauturu-o-Toi/LBI (2817 ha, S36° 11' 32, E175° 4' 29) is a rugged volcanic island, also in the Hauraki Gulf region (Figure 1.1). Although habitat loss in the form of logging of kauri took place prior to the sanctuary being declared in the 1890’s (Campbell 2011), the island is predominantly unmodified primary forest and is considered one of the most intact native ecosystems in New Zealand. It is managed as a closed scientific sanctuary by DOC, with no public access. A remnant wild population of RCP is present on the island, which was declared free of exotic predators in 2004 following the eradication of the kiore (Campbell 2011). The island is supported by the
Hauturu-o-Toi/LBI Trust, and represented by iwi including Ngati Wai, Ngati Manuhiri, and Ngati Rehua.
Factors for disease transmission
There are aspects of each island that must be incorporated into interpretation of results from health studies, particularly when surveying for diseases of parrots.
Sympatric parrots are present on both islands, however Tiritiri Matangi Island only hosts Eastern Rosella (Platycercus eximius), with the occasional non-resident North
Island kākā (Nestor meriodionalis) (Graham et al. 2013). Hauturu-o-Toi/LBI is home to North Island kākā, Yellow-crowned Parakeets (Cyanoramphus auriceps) and, since
2012, the critically endangered flightless kākāpō. Of these parrot species, kākā and
Eastern Rosella are the only species likely to move between the Hauraki Gulf islands and the mainland. RCP have also been recorded to travel ca.65km (Ortiz-Catedral
2010), therefore they have the potential to move to mainland or other offshore
12 sites. Migrating birds have been shown to be a source of pathogen spread between populations, with implications for emergent diseases (Lawson et al. 2011). There are management activities on Tiritiri Matangi Island that may interact with disease transmission dynamics, such as the use of communal feed and water stations
(Rimmer 2009). Feed stations are for the nectar-feeding hihi (Notiomystis cincta) and
Bellbirds (Anthornis melanura), and exclude non-target species. Water points are available in selected areas of the island and all birds may access them. These supplementary programs may alter disease transmission by artificially boosting populations (increasing contact rates and population densities) and encouraging congregation within and between species (Robinson et al. 2010). To combat some of these potential effects, a program is in place to clean feeders and water troughs regularly, however this will not prevent environmental contamination particularly of heavily used perches adjacent to feed and water stations, nor feed and water contamination between cleaning sessions.
1.4 Diseases of wild Red-crowned Parakeets and sympatric parrot species
1.4.1 Infectious diseases and parasites detected in New Zealand
There have been substantial investigations into the diseases of free-ranging parrots in New Zealand, partly driven by policy from DOC on health and disease screening prior to movements, as well as a general awareness amongst conservation biologists in the region as to the risks posed to small populations by disease. BFDV is by far the most researched infectious disease of wild parrots in New Zealand and globally,
13 reflecting the genuine threat posed by this virus for parrot conservation (Kundu et al.
2012, Massaro et al. 2012, Peters et al. 2014). It has been considered a conservation threat in the Mauritius Parakeet (Psittacula echo) in Mauritius (Kundu et al. 2012), the Orange-bellied Parrot (Neophema chrysogaster) in Australia (Peters et al. 2014), and the Cape Parrot (Poicephalus robustus) in South Africa (Regnard et al. 2014b).
However, there may also be a component of research myopia, leading to a tendency to screen for known pathogens rather than systematically review all potential pathogens. BFDV has been detected in wild and captive parrots (native or introduced) in New Zealand, including RCP on the North Island (Ortiz-Catedral et al.
2009b), Yellow-crowned Parakeets on the South Island (Massaro et al. 2012), and introduced Eastern Rosellas and Sulphur-crested Cockatoos (Cacatua galerita) on the
North Island (Ha et al. 2007, Massaro et al. 2012). Surveillance has failed to detect the virus in kākā, Malherbe’s Parakeets, kea (Nestor notabilis), kākāpō, or Forbes
Parakeets (Cyanoramphus forbesi) (Massaro et al. 2012), however sample sizes in the kākā and kea were insufficient to give statistical confidence that the population is free of the pathogen at a low prevalence regionally. The virus causes highest mortality in nestling and young birds (Schoemaker et al. 2000, Doneley 2003, Jackson et al. 2014a), with older birds recovering, becoming carriers, or succumbing to chronic disease characterised by feather dystrophy and immunosuppression (Raidal
1995). The environmental stability of the virus (Todd 2000), the potential for non- clinical carriers or reservoirs (Jackson et al. 2014b), and the excretion of high viral loads in feather and faeces of affected birds (Raidal et al. 1993b) are three key epidemiological factors that enable persistence and spread within populations.
14 There is limited data on the parasites of wild RCP and sympatric parrots in New
Zealand. Hippoboscids have been described from wild RCP, including hyperparasitism of the fly with the skin mite Promyialges macdonaldii (Heath 2010).
Both the fly and the mite are cosmopolitan species with a global presence, and the mite has been reported to cause mange in avian species elsewhere (Gilardi et al.
2001).
1.4.2 Disease surveillance of RCP on Tiritiri Matangi Island and Hauturu-o-Toi/LBI
RCP have been extensively surveyed for diseases on Tiritiri Matangi Island and
Hauturu-o-Toi/LBI, largely due to research activities and the translocation potential of the remnant wild Hauturu-o-Toi/LBI population. In 2002 as part of a Ministry of
Agriculture and Fisheries (now MPI) biosecurity project, RCP on Tiritiri Matangi Island were sampled and found to be negative for Salmonella sp. (by culture, n=53), BFDV
(by PCR, n=67), avian influenza (by culture, n=49) and avian paramyxovirus (by culture, n=48) (Jakob-Hoff 2002). Further surveillance of RCP from Tiritiri Matangi
Island, Hauturu-o-Toi/LBI, and remote Raoul Island for Salmonella sp. and Yersinia sp. (by culture, n=101) and Campylobacter sp. (by culture, n=82) did not yield any positive results (Ortiz-Catedral et al. 2009a).
The first detection of BFDV in a wild native parrot in New Zealand was on Hauturu-o-
Toi/LBI during a translocation trip in 2008, with molecular evidence suggesting a recent outbreak or spill over (Ortiz-Catedral et al. 2010). Infected birds presented with loss of wing and tail feathers, as well as body contour feathers, leading to an inability to fly in some individuals (Ortiz-Catedral et al. 2009b).
15 Plasmodium relictum (avian malaria) has also been detected in RCP on Hauturu-o-
Toi/LBI, at a prevalence of 40.9% (95%CI: 20.49-61.51%), without evidence of clinical signs or haematological changes at the time of sampling (Ortiz-Catedral et al. 2011).
Plasmodium species have been detected in a range of mostly passerine species across New Zealand (Howe et al. 2012). Although occasional mortalities have been reported in wild native species in New Zealand (Howe et al. 2012), to date no deaths have been reported in native or exotic parrots as a result of Plasmodium infection.
Reported ectoparasites of free-living RCP on Hauturu-o-Toi/LBI include an unidentified mite (possibly Foriculoecus spp., Protalges spp. or Pseudoallophinus spp.), that was found in feather follicles during post-mortem examination of birds that had feather loss and BFDV infection during a translocation event (Brennan and
Alley 2008). Interestingly, these unidentified mites were found in 7/7 specimens, however histopathological evidence of BFDV infection was only evident in 2/7 birds examined.
1.5 Disease implications of aviculture and parrot trade in New
Zealand for wild RCP
1.5.1 Trade and the establishment of wild exotic parrots in New Zealand
Early records of released captive Eastern Rosellas in the Dunedin region in 1910
(Wright and Clout 2001) indicate trade and movement of exotic parrots from
Australia to New Zealand had been occurring long before biosecurity or conservation implications were considered. Historical releases (both deliberate and accidental) have enabled Eastern Rosellas and Sulphur-crested Cockatoos to establish across the
16 majority of the North Island, although notably Eastern Rosellas failed to expand their range on the South Island (Wright and Clout 2001). The deliberate release and subsequent establishment of Rainbow Lorikeets (Trichoglossus haematodus) in the
Auckland region in the 1990’s led to their listing as an Unwanted Organism under the
Biosecurity Act 1993 (http://www.doc.govt.nz/conservation/threats-and- impacts/animal-pests/animal-pests-a-z/rainbow-lorikeet/). An eradication program led by DOC, MPI and regional councils effectively removed the threat, although require annual surveillance and targeted captures to counter ongoing accidental and deliberate releases of this species (Polkanov 2006-2009).
The domestic parrot population is un-quantified in New Zealand, however a wide variety of parrots are available through the local trade (Holden 1997). Importation of Australian parrots ceased in the 1960’s following the Australian ban on export of native birds except for scientific or zoological reasons (Holden 1997). In 1997 New
Zealand placed a ban on all parrot imports for biosecurity reasons to protect the native parrot fauna. However it is still legal to export exotic parrots (including native
Australian parrots) from New Zealand. This, in combination with legal loopholes preventing effective prosecution, has led to the belief that New Zealand is being used as a laundering point to legally export endemic Australian birds, the majority of which are parrots (Holden 1997). Recent reports indicate around 600 Convention on
International Trade on Endangered Species (CITES) listed birds (the majority being endemic Australian parrots) are legally exported annually from New Zealand (Holden
1997). The ‘species traded’ list includes some that are known to be difficult to breed in captivity, thus arousing suspicion as to their provenance. Prosecutions led by
17 enforcement agencies from the USA, Australia and New Zealand, have identified highly organised smuggling rings operating across these countries to smuggle birds via light aircraft from the east coast of Australia to New Zealand (TRAFFIC 2011).
Clearly there is an illegal trade operating to bring parrots into New Zealand, with disease risk implications for the conservation of wild native parrots through biosecurity breaches.
1.5.2 Diseases of domestic parrots in New Zealand
Targeted disease surveillance in the domestic parrot population has been limited in
New Zealand. BFDV was reported in eight species of captive exotic parrots in a study to investigate host specificity of the virus (Ritchie et al. 2003). Two Rosellas (species unspecified) taken to a domestic veterinarian tested positive in 2002 for
Psittacinepox virus, an exotic unwanted organism in New Zealand (Gartrell et al.
2003). This led to an MPI investigation of several aviculture properties to contain the outbreak. Knemidokoptinid mange (species unspecified) was incidentally detected in wild-caught but captively-held Eastern Rosellas during the investigation.
Knemidocoptes pilae has been reported from captive Budgerigars (Melopsittacus undulates) (O'Grady 1960), and Yellow-crowned Parakeets (Bishop and Heath
1998a). Surveillance for multiple pathogens across three breeding aviaries of budgerigars detected a novel Polyomavirus and Plasmodium sp., however did not detect BFDV (Baron et al. 2014).
18 1.5.3 Disease implications of the parrot trade in New Zealand
The global trade in wild or exotic species is a considerable disease threat for native wildlife and ecosystems (Karesh et al. 2005, Travis et al. 2011). Some consider that translocations for the pet or wildlife trade constitute an equal or greater risk for disease introduction, as translocations for conservation outcomes (Kock et al. 2010).
The statistics on the legal trade in live birds alone are confronting, with estimates of
1.5 million birds traded annually based on customs and CITES data (Roe et al. 2002).
However this figure underestimates the scale of trade, due to the unquantifiable illegal trade in wildlife, considered commercially second only to the illegal trade in narcotics and arms (Roth and Merz 1997). The illegal trade in parrots is driven by high prices and demand for species that are rare or difficult to obtain, or subject to trade restrictions imposed by various national policies. Of the 36% of parrots that are listed in a threatened category by the IUCN, over half of these are reported to be threatened by trade (Pain et al. 2006).
Trafficking of parrots, legal or illegal, has significant implications for conservation managers worldwide beyond population impacts at the source site, due to the risks of artificially expanding a pathogen’s host or geographic range. The illegal trade, by its nature, is not subject to quarantine measures or disease screening of individuals prior to, or at the point of, movements. Transport conditions and holding facilities may encourage the expression and shedding of disease agents as a result of stress, making this group potentially higher risk than legally traded parrots. The number of reported pathogens in domesticated parrots substantially outweighs those known from their wild counterparts, which may partly reflect the higher likelihood of
19 disease detection in captive birds versus wild birds, including limitations of effective wild bird surveillance. Global movements of pathogens via the trade in birds have been implicated in historical species and range expansions of viruses such as BFDV
(Varsani et al. 2011 ). The threat posed by the legal and illegal trade in birds does not relate solely to pathogen spread. Captive conditions provide opportunity for pathogen evolution due to mixing of host species and pathogen strains, with potential consequences for enhancement of virulence, pathogenicity and cross- species transmission (Julian et al. 2013, Sarker et al. 2014c).
New Zealand is considered free of several potentially fatal infectious diseases known from captive parrots elsewhere including Psittacid herpesvirus-1 and Psittacinepox virus, thus there is a genuine need to maintain strict biosecurity measures to prevent these exotic diseases entering the country via trade. It should be noted however that there is no available commercial diagnostic test in New Zealand for the two viruses mentioned, thus it may be a case of absence of evidence rather than evidence of absence, and the reliance is on pathologists and veterinarians to refer suspicious cases to MPI for further screening (Gartrell et al. 2003). Release pathways for pathogen spread from captivity to the wild in New Zealand include accidental and deliberate releases of parrots carrying disease-causing agents, including the use of captive facilities for breeding and release programs. As discussed, all these situations have occurred in New Zealand, and continue to occur, so the risk is real and requires ongoing management. Appropriate quarantine, health screening, and risk management protocols can be put in place to ensure breed for release programs mitigate the risks of releasing pathogens at destination sites (Jackson et al. 2014a).
20 1.5.4 Disease risk management for conservation
Mitigating disease risks in threatened species programs has become easier with the development of new tools and decision-making methods to guide management decisions (Jakob-Hoff et al. 2014). These sorts of tools are critical in translocation programs, which frequently have high levels of uncertainty. A commonly noted issue in disease risk analyses is the paucity of data from which to make decisions, as well as benchmark or interpret results (Makan 2009). Although decision-making tools assist the end user in navigating uncertainty, it is still crucial to gather useful data where possible to feed into these tools. Invariably the quality of the output reflects the information or data it is based upon.
1.6 Aims of this study
The primary aims of this study were based on the need to investigate reported feather loss in the Tiritiri Matangi Island RCP population and to generate baseline normal health data for this population.
Therefore the study aimed to:
Determine baseline normal physical, haematology, and biochemistry reference
ranges for the species, including creatine kinase (CK).
Determine BFDV presence, prevalence, seasonal trends and conduct risk factor
analysis.
Investigate feather loss including prevalence, potential causes, pathology,
seasonal trends and conduct risk factor analysis.
21 Investigate breeding success of RCP affected by feather loss, including health of
chicks fledging from affected and unaffected nests.
This information will be used to guide future disease risk analyses for this species, to expand our understanding of causes of feather loss in wild parrots, and to provide baseline information for organisations involved in the captive and wild care of this and closely related species.
22
CHAPTER 2:
GENERAL METHODS
23 2.1 Mist netting
Birds were captured using standard 3m high mist nets with 30mm mesh (Avinet,
USA), of variable lengths (6m, 9m and 12m). Nets were erected on poles up to 5m high (Figure 2.1), depending on the height of adjacent habitat or flight paths of the birds. Mist nets were operated from dawn until dusk, and closed if conditions were unsuitable from a welfare perspective (e.g. rain or high temperatures). Mist nets were observed continuously from a distance to retrieve birds as soon as possible after capture.
Figure 2.1: Mist net 9m x 3m erected on 5m high poles and placed in flight path at
Emergency Landing, Tiritiri Matangi Island.
2.2 Locations of captures
Sites for captures were chosen based on the following criteria:
High likelihood of capture of the target species
Representation of different areas of the island
Lower numbers of non-target species
24 The sites chosen for captures on Tiritiri Matangi Island are shown in Figure 2.2.
Figure 2.2 – Map showing sites of captures and sampling of Red-crowned Parakeets on Tiritiri Matangi Island.
Due to the geography of Hauturu-o-Toi/Little Barrier Island (LBI), as well as the high canopy and low likelihood of RCP captures deeper into forest, all captures took place in the edge habitat and open flats adjacent to the ranger station and bunkhouse
(Figure 2.3).
25
Figure 2.3: Map showing site of captures and sampling of Red-crowned Parakeets on
Hauturu-o-Toi/Little Barrier Island.
2.3 Handling and removal
Birds were removed from mist nets by experienced handlers, placed into calico bags measuring 25cm x 35cm (Prospectors Earth Sciences, Australia), and taken to the processing station. Birds that were unable to be processed immediately were placed into a plastic bird carrier (My Pet Mobile, PetMart, Auckland, NZ), and kept in a shaded quiet area adjacent to the processing station.
2.4 Bycatch
Non-target species were released immediately, with details of time of capture, sex if known, bands if present, and any comments recorded on an incidentals sheet.
26 2.5 Anaesthesia and sample collection
2.5.1 Set up of field anaesthetic station
A field processing station was set up near to mist netting, either in the field if remote from buildings, or in a suitable building if available. The station consisted of a plastic trestle table, with anaesthetic and sampling equipment at one end, and sample pots and data files at the other end (Figure 2.4). This enabled a separation of ‘dirty’ equipment (used for sampling birds) from ‘clean’ equipment (sample pots, labels, data sheets), to prevent contamination of materials that would go to the laboratory.
See section 2.7 for details of the disinfection protocol used between birds.
Figure 2.4: Sampling a Red-crowned Parakeet on Tiritiri Matangi Island. In the foreground an RCP can be seen under anaesthesia, with a nurse monitoring vital signs and the author to the left of the image. The end of the table in the foreground is considered dirty and thoroughly cleaned and disinfected between sampling birds,
27 the far end of the table is considered clean with samples and data management being conducted by Auckland Zoo keeping staff.
2.5.2 Roles of personnel
To prevent contamination of sample pots particularly with BFDV, roles were assigned to ensure any person handling birds did not touch sample pots and other ‘clean’ equipment.
Person 1 (the veterinarian): considered ‘dirty’ and did not touch anything considered
‘clean’ e.g. sample pots, writing equipment, storage containers for equipment, until hands were cleaned with Trigene® (Ethical agents, Auckland, New Zealand) or F10.
Responsible for:
Removing the bird from the net
Anaesthesia of bird
Collection of all samples
Physical examination of bird
Recovery of bird
Release of bird
Person 2 (experienced bird handler/mist netter): also considered ‘dirty’ and did not touch anything as per person 1.
Responsible for:
Removing the bird from the net
Assistance during recovery and release of bird
28 Person 3 (experienced bird anaesthetist): also considered ‘dirty’ and did not touch anything as per person 1.
Responsible for:
Anaesthetising and monitoring the bird during anaesthetic
Recovery of bird
Person 4 (data collector): considered ‘clean’, therefore must not touch the bird or anything that has come in contact with the bird.
Responsible for:
Recording all information on the data sheet (Appendix 1)
Recording the procedure including time of capture, time of removal from mist net, time of anaesthetic start/finish, time of recovery (awake), time of release
Presenting sample jars/pots etc to the veterinarian, ensuring they are labelled appropriately, and that they are stored away from the site of anaesthesia so they do not become contaminated
Taking photographs as requested
2.5.3 Anaesthetic protocol
All birds were anaesthetised using a portable field anaesthetic machine (Stinger,
Advanced Anaesthesia Specialists, Sydney, Australia), with isoflurane delivered via a zero dead space anaesthetic mask (AAS, Sydney, Australia) designed for small birds and mammals. Birds were restrained in the calico capture bags, their heads exposed and a mask fitted for the induction of general anaesthetic. Isoflurane was initially
29 delivered at 5% for approximately 30 seconds or until relaxation occurred, reduced to 2-3% for a further 30 seconds or until the withdrawal reflex and wing tone were reduced, then generally maintained at 1-1.5% with an oxygen flow rate of
400ml/min for the duration of anaesthesia, which lasted an average of 12 minutes.
At the end of the procedure birds were given oxygen until first responses were observed (head movement, opening eyes, attempts to move). Once birds were considered recovered (head control, eyes open and responsive to environment and stimulus), they were placed in the plastic bird carriers until assessed as ready for release.
During anaesthesia all birds were provided with 0.9% NaCl subcutaneously in the inguinal region at 0.5-1% of body weight (i.e. 0.5-1.0mls). Supplementary heat in the form of heat pads was also provided unless ambient temperatures precluded the need. Heart rate and respiratory rate were recorded using a paediatric stethoscope, and cloacal temperature was recorded. These measurements were made at the start of each anaesthetic, and monitored during anaesthesia by the avian anaesthetist.
2.5.4 Sample collection
Blood
Following induction of anaesthesia and collection of baseline physiological data, blood was obtained from the medial metatarsal vein using a 25-gauge ¾ inch needle attached to a 1ml syringe. Approximately 0.1-0.2mls was obtained, except in a sub- set of birds (n=36) which were bled by jugular venipuncture to obtain sufficient
30 blood for biochemistry (maximum 0.5mls). These latter birds were sequentially sampled in the first field session, and the maximum volume did not exceed the safe limit of 1% of body weight reported for blood sampling in birds. Cotton swabs were applied with pressure following blood taking to assist haemostasis.
Blood was divided in the following way:
1-2 drops in Longmire’s lysis buffer and stored at room temperature
2 drops on filter paper (Whatmann’s No.3 insert) and stored at room temperature
2 blood smears made on slides with frosted end, using a designated glass spreader slide with bevelled edges, air-dried and stored at room temperature
1 microhaematocrit tube sealed and stored upright
Remaining blood (if available) placed in paediatric serum gel separator tubes for biochemistry
Feathers
Four feathers were plucked from the breast area by grasping near the base gently with forceps and removing in the direction of feather growth. Where possible blood or growing feathers were collected preferentially. Two feathers were placed into
1.5ml eppendorf tubes filled with 70% ethanol, while the other two were placed into an envelope or ziplock bag for storage.
Skin biopsy
In the second year of the study (2012), skin biopsies were collected from all birds regardless of feather or skin condition. The site chosen was the back of the
31 head/neck, just caudal to the base of the skull. Further details on the selection of site are provided in Chapter 5. A 3-5mm piece of skin, with growing feathers where possible, was cut using fine sharp scissors, and placed in a sample pot containing
10% neutral buffered formalin. Pressure was briefly applied to the site with a gauze swab, then a single suture placed using 4-0 absorbable polyfilament suture material
(Visorb, SVS, Auckland). A small dab of betadine ointment was applied using a cotton tipped swab.
External parasites
No systematic protocol was established to recover ectoparasites, except where opportunistically observed. Hippoboscid flies were collected into 1.5ml eppendorf tubes with 70% ethanol when caught whilst handling birds.
2.6 Physical examination, morphometric measurements and photographs
A full physical examination was performed on all birds under general anaesthesia, including body condition, skin and feather condition, eye exam, oral exam, palpation of the coelomic cavity, cloacal exam, and auscultation of the lungs and air sacs.
Morphometric measurements were taken using vernier callipers to record the culmen width and length, flattened wing chord, and tarsometarsus. All birds were banded with 3 D-size colour butt bands (DOC, Wellington, New Zealand) and 1 D-size metal band (DOC, Wellington, New Zealand) on the tarsometatarsus. Metal bands were only used as the bottom band in the designated combination, to prevent
32 slipping over colour bands. Metal bands were positioned and tightened using banding pliers (Department of Conservation, Wellington, New Zealand).
Subjective body condition scores were assigned by assessing muscling over the keel as reported for black cockatoos (Calyptorhynchus spp.) (Le Souef 2012), as well as palpating subcutaneous fat stores. The author observed and recorded body condition for all birds. The four categories were “poor”, “average”, “good”, and “very good”.
Birds were weighed using digital scales (Wedderburn, Auckland, New Zealand). A standard set of photographs was taken with a laminated re-usable label showing the birds ID, date and location (Figure 2.5), including:
a. Ventrum with wings extended
b. Dorsum with wings extended
c. Head photos (from each side and from the front)
33
Figure 2.5: Examples of standardised photographs obtained of Red-crowned
Parakeets during the study.
2.7 Disinfection/biosecurity protocols
All calico bags that had held birds were soaked in Trigene® (Ethical agents, Auckland,
New Zealand) at a dilution of 1:20 overnight, then washed and dried before being used again. Plastic bird carriers were sprayed with Trigene® (1:100) until wet, left for a minimum of 30 mins prior to being wiped clean and re-used.
Plastic gloves were worn by the veterinarian and anaesthetist during the procedure, to protect from zoonotic disease and also to reduce the risk of disease spread between individuals. All equipment including the face-mask, anaesthetic machine, trestle table surface, sampling and measuring equipment were sprayed with
Trigene® (1:100) and left in contact with the disinfectant until the next bird
34 (minimum 5 minutes). All instruments were soaked in Trigene® (1:100) then rinsed and dried between sampling. Washable covers were used for heat pads, replaced between birds, and soaked/washed overnight as per the calico bags.
2.8 Animal ethics
Animal ethics approval for all work conducted was provided through the Murdoch
Animal Ethics Committee (W2390/11) and the Auckland Zoo Animal Ethics
Committee.
35 Chapter 3:
Haematology, biochemistry and body
weight reference ranges for wild Red-
crowned Parakeets
This chapter is modified from the following paper prepared for submission:
Jackson, B., Jakob-Hoff, R., Holyoake, C., Robertson, I., Varsani, A. and K.Warren.
2014. Haematology, biochemistry, and physical data reference ranges for wild Red- crowned Parakeets (Cyanoramphus novaezelandiae) of New Zealand, including a comparison of sexing techniques.
36 3.1 Introduction
Reference ranges for health parameters of wildlife species are necessary for a range of purposes including assessment of population health (Mathews et al. 2006), interpretation of diagnostic tests used for screening prior to translocations, and captive management of threatened species including those intended for release to the wild (Polo et al. 1998). Marked differences may exist in biochemical and haematological reference ranges even within the same avian family (Polo et al. 1998,
Briscoe et al. 2010), highlighting the need for species-specific references ranges (Low et al. 2006). Basic data that complements health assessments of avian species includes knowledge of the normal body weight and morphometric measurements, as well as the ability to accurately determine sex, whether by physical characteristics or
DNA-based molecular sexing techniques.
Haematology and serum biochemistry are standard diagnostic tools used as an adjunct to physical examination to infer the health status of birds in captivity
(Briscoe et al. 2010). Total (or estimated) and differential white cell counts are routinely collected to look for evidence of inflammatory or infectious processes
(Briscoe et al. 2010), although compared to mammalian haematology, there are knowledge gaps regarding the function of certain avian blood cells that affect interpretation of the haemogram (Mitchell and Johns 2008). Biochemistry provides information on the health of target organs such as the kidney and liver, as well as general metabolic states (Harrison and Lightfoot 2006). In wild species, monitoring the enzyme creatine kinase in the biochemical profile can be used to indicate the level of muscle damage in an individual, and can indicate the development of
37 capture myopathy, a potentially fatal condition that may arise from capture and restraint (Ward et al. 2011). Importantly, whilst biochemistry and haematology are useful tools in the assessment of health status, changes in these parameters are not always evident in a diseased bird (Briscoe et al. 2010), and test results should therefore be interpreted with caution and assessed in combination with other diagnostic tools such as the complete physical examination and appropriate diagnostic screening for targeted pathogens or toxins.
Barriers to creating reference ranges for health parameters in wild avian species include the cost of sampling and diagnostics (McDonald et al. 2010), the size and type of sample required, logistics of sample storage and transport in remote areas, and conflict between cheap or transportable methods that may be applied in the field versus gold standard laboratory-based diagnostics. Although using captive birds to generate reference ranges may offer a solution to some of these problems, management and husbandry factors can lead to notable differences in haematology and biochemistry results compared with wild populations (McDonald et al. 2010), meaning some caution must be applied when interpreting wild results against data derived from captivity. Intrinsic (e.g. age, sex, reproductive status and stage of moult), and extrinsic (e.g. environmental impact on resources) factors may affect results of diagnostic assays and influence the validity of reference ranges or interpretation of results (Capitelli and Crosta 2013). Thus collecting samples in a manner that reflects wild conditions in the species of interest, and reporting results with these limitations in mind remains key to providing appropriate baseline data.
38 Red-crowned parakeets (RCP, Cyanoramphus novaezelandiae) belong to the
Cyanoramphus genus, a group that has suffered substantial population declines or range restrictions in New Zealand (Boon et al. 2001), and includes the critically endangered Malherbe’s Parakeet (Cyanoramphus malherbi). RCP are currently restricted to predator free or controlled islands and mainland sites (Boon et al. 2001) and, as a species, are commonly used in translocations due to their relatively flexible life history traits and capacity to adapt to modified environments (Ortiz-Catedral and
Brunton 2010b). Future movements and/or captive breeding programs will require reliable species-specific data from which to infer the health status of individuals and populations for conservation management.
The aim of this study was to determine baseline reference ranges for haematology, biochemistry, and body weight in healthy wild RCP, to be used when dealing with wild captured birds for captive management or for reintroductions/translocations. A secondary aim was to examine creatine kinase concentration in a select group of birds to infer the impact of capture methods on this species. Finally, we assessed agreement between DNA sexing and morphometrics as these are commonly used to determine the sex of this species.
3.2 Methods
Capture sites and sampling
Birds were captured in standard 30mm diameter mist nets (Avinet, USA) between
2011-2013. Capture sites were on Tiritiri Matangi Island (TTM, S36˚ 36' 2", E174˚ 53'
24") and on Hauturu-o-Toi/Little Barrier Island (LBI, S36˚ 12' 3", E175˚ 4' 54.84"),
39 both predator-free island sanctuaries located in the Hauraki Gulf region of Auckland,
New Zealand. All birds were placed in single use calico bags (Prospectors Earth
Sciences, Australia) and taken to a nearby field station for processing. Individuals were anaesthetised using a portable anaesthetic machine (AAS, Australia) with a zero dead space circuit designed for small mammals and birds (AAS, Australia).
Induction was with 5% isoflurane in oxygen for approximately 30 seconds, followed by maintenance at 1-2%. All birds were given 0.5mls of subcutaneous 0.9% NaCl, and warmth was provided when appropriate using wheat bags and single use fleece pouches. Following induction and assessment of heart rate, respiratory rate and body temperature, approximately 0.1-0.2mls of blood was obtained from the medial metatarsal vein using a 25-27g needle and 1ml syringe. This was placed in a micro haematocrit tube and used to make two air-dried blood smears for haematology, with the remainder placed on filter paper, and in lysis buffer for studies not reported here. Blood smears were made using glass slides pre-cleaned with ethanol, and a dedicated slide spreader to minimise damage to blood cells and improve the quality of smear (Capitelli and Crosta 2013). Smears were submitted to the laboratory at the end of each field session. For a subset of birds (n=36) approximately 0.4mls of blood was collected by jugular venipuncture and placed in paediatric serum gel tubes to process for biochemistry. These birds were sampled in sequence (i.e. the first 36 captured) from the first field session, to reduce selection bias. Serum was separated nightly by centrifugation for 10 mins at 2500rpm (Zip Spin ZS-1, LW Scientific, USA), stored in a refrigerator, and processed within 4 days. All birds had a full physical examination including auscultation of the air sacs and lungs, and palpation of the coelomic cavity.
40 Haematology and biochemistry
Blood smears were fixed in methanol, then placed for 1 minute in a bath of
Leishman’s stain, pH 6.8 phosphate buffer, and May Grunwald stain diluted 50:50 with pH 6.8 phosphate buffer (Dacie and Lewis 1995). Estimated white cell counts
(EWCC) were calculated by counting the number of white blood cells (WBC) in 10 fields at 40x objective, using an area where red blood cells (RBC) formed a uniform monolayer on the slide. This number was divided by 10 to get an average and multiplied by 2 to obtain the EWCC x 109/l (Samour 2000). Differential counts were performed using oil immersion at 50x magnification. Each WBC was recorded until
100 WBC were counted, to obtain a relative differential percentage. The absolute differential was calculated by multiplying the relative differential (%) x the EWCC, to obtain a count x 109/l. Examination of smears was performed by experienced avian haematologists at Gribbles Veterinary Pathology, Auckland, for samples in 2011, and at New Zealand Veterinary Pathology, Auckland, for samples in 2012-13.
A Roche/Hitachi Modular P Chemistry Analyzer (Roche, NZ) was used to measure creatine kinase (CK), bile acids (BA) and uric acid (UA) in serum samples. The time from capture to anaesthesia was used as proxy for the time at which CK was measured, given blood samples were obtained from all birds immediately after induction. CK continues to be released from muscles post-capture (Bailey et al. 1997) and should be reported against time, as samples obtained soon after capture may appear falsely low. Therefore results are presented to reflect the degree of change over time.
41 Morphometrics
Culmen width/length and the tarsometatarsus length were measured using manual
Vernier calipers accurate to the nearest 0.1mm, and wing chord was measured with a stopped steel ruler accurate to the nearest 1.0mm. Weight was recorded with digital scales (Wedderburn, NZ) accurate to 1g.
Statistical analyses
All analyses were performed in “R studio” version 0.98.501 (Ihaka and Gentleman
1996). For haematology, biochemistry and body weight reference ranges, the entire dataset was examined and all birds with clinical signs that may be attributed to ill health including feather loss or poor body condition were removed (n=55). The remaining dataset was checked for normality using the Shapiro-Wilks test, with tests for significance using the two-tailed t-test for normal data, or the Kruskal-Wallis rank-sum test for non-parametric data. For morphometric reference ranges the entire dataset was used, and analyses were performed as for the haematology and biochemistry parameters.
Molecular analyses
Total DNA was extracted from feather or blood samples using the iGenomic blood
DNA extraction kit (Intron Biotechnology, Korea) following the manufacturers protocols. Four microlitres of the extracted DNA was used as a template for sex determination by CHD-gene targeted PCR using primer pairs 2550F (5´-GTT ACT GAT
TCG TCT ACG AGA-3´) and 2718R (5´-ATT GAA ATG ATC CAG TGC TTG-3´)(Fridolfsson and Ellegren 1999). The following thermal cycling conditions were used: initial
42 denaturation of 95°C for 3 min; 35 cycles of 95°C for 30s, 50°C for 30s, and 72°C for
45s; and a final extension of 72°C for 1 min. The amplicons were separated on a 2% agarose gel stained with SYBR® Safe DNA Gel Stain (Life technologies, USA) and visualised to determine the sex.
Following comparison of the results of DNA sexing with morphometric data, results were confirmed by a repeated CHD-gene targeted PCR assay of any birds that did not show agreement between DNA sexing and morphometric analysis (n=6).
3.3 Results
Haematology and biochemistry
A total of 229 birds were captured during the sampling period, with 174 remaining for analyses once birds with abnormal clinical signs were removed. Not all parameters were available for all birds therefore sample sizes vary for specific parameters. Recorded values for UA, BA, and PCV followed a normal distribution, whilst all other haematological and biochemical parameters had a non-parametric distribution. No significant differences were found between all parameters based on sex therefore reference ranges are presented with both sexes combined. A summary of the results for haematology and biochemistry assays is provided in Table 3.1.
Twelve individuals had CK levels >800 U/L on first analysis; however, due to insufficient serum a dilution was not possible to obtain an accurate value, and therefore these are not reported.
43 Table 3.1: Selected haematology and serum biochemistry reference ranges for Red- crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-Toi/Little Barrier Island,
New Zealand.
Parameter n mean median sd 10-90 range percentile
PCV (%) 100 55 55 5 50-60 42-67
TP g/l 80 38 38 6.6 30-47.1 25-57
EWCC x 109/L 162 8.3 7.5 4.3 4-14.38 2-24
Heterophil x 109/L 162 2.0 1.6 1.5 0.5-3.8 0.2-8.3
Heterophil (%) 162 24.1 21 13.2 8.1-42.9 3-66
Lymphocyte x 109/L 162 5.3 4.1 3.3 2.01-10.2 1.1-16
Lymphocyte (%) 162 62.1 65 17.1 35-82 15-91
Monocyte x 109/L 152 0.53 0.4 0.6 0.1-1.38 0-4.1
Monocyte (%) 152 7 6 6.9 1-16.9 1-45
Eosinophil x 109/L 151 0.5 0.4 0.6 0.1-1.1 0-3.4
Eosinophil (%) 151 5.3 5 4.4 1-12 1-26
Basophil x 109/L 106 0.1 0.1 0.2 0.0-0.4 0-0.8
Basophil (%) 106 1.4 2 1.7 1-5 1-9
Uric Acid µmol/l 28 532.8 423 272.45 220-881 120-1090
Bile Acids µmol/l 16 24.18 21.5 11.58 12-37.5 6.2-43
Creatine Kinase IU/l 16 2123 756 2424 515-5945 387-7597
Correlation between the CK measure and the time to sampling (i.e. the time from capture to obtaining a blood sample) is shown in Figure 3.1. A linear relationship was established between this enzyme and the time from capture to sampling (Pearson’s correlation co-efficient = 0.95, 95%CI: 0.84-0.98, t = 10.43, df = 13, p<0.0001).
44
Figure 3.1: Correlation between creatine kinase levels (U/l) and time from capture to sampling (min) in wild Red-crowned Parakeets on Tiritiri Matangi Island, New
Zealand.
Morphometric measurements and bodyweights
Descriptive statistics for standard morphometric measurements (mm), including those used to determine sex in RCP (culmen width and length), are presented by sex in Table 3.2. Body weights for different locations are presented by sex in Table 3.3.
Body weight and morphometric measurements for both sexes followed a normal distribution, although weight was non-parametric for both sexes prior to removing birds considered abnormal due to feather loss or other clinical signs.
45 Table 3.2: Descriptive statistics by sex for standard morphometric measurements
(mm) from normal Red-crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-
Toi/Little Barrier Island.
Measurement (mm) Sex Mean sd Range n
Culmen width M 10.62 0.42 9.4-11.7 117
F 9.25 0.37 8.4-10.2 111
Culmen length M 16.81 1.39 11.4-19 93
F 13.57 0.72 11.5-16 82
Wing chord M 134.5 3.98 125-145 116
F 127 3.82 117-137 111
Tarsometarsus M 22.18 0.84 18.8-24.1 117
F 21.0 0.86 19.1-23.5 111
Table 3.3: Descriptive statistics by sex and location for body weights (gm) of normal
Red-crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-Toi/Little Barrier
Island.
Location Sex Mean sd Range n
Tiritiri Matangi Island M 87.68 6.26 75-106 65
F 67.15 5.65 53-86 65
Hauturu-o-Toi/LBI M 91.0 8.53 71-106 24
F 71.0 5.83 62-83 21
Females on Hauturu-o-Toi/LBI had significantly higher mean weights compared to those on Tiritiri Matangi Island (p=0.013), however the difference between males was not significant (p=0.09). Seasonal differences were not detected in this study when comparing male weights between April and September on Tiritiri Matangi
46 Island (p=0.61) or female weights between April and September on Tiritiri Matangi
Island (p=0.14).
Morphometric measurements and DNA sexing data
Comparison of the results of DNA-based sexing assays and measurements of culmen width and length are summarized in Figure 3.2. Generally there was agreement between morphometric measurements of the beak and DNA-based sexing, however, three birds were found to be male via DNA sexing with beak measurements that clustered with females based on both culmen length and width. A further three birds had culmen lengths that aligned with females, however they were within the width range for males, and were DNA-sexed as males. Body weights of these six birds aligned with normal male body weights.
Figure 3.2: Comparison of beak measurements (culmen width and length in mm) with PCR-based DNA sexing results for Red-crowned Parakeets on Tiritiri Matangi
47 Island and Hauturu-o-Toi/LBI. Individuals sexed as males by PCR are indicated by white dots, females by grey dots.
3.4 Discussion
This study provides the first biochemical and haematological reference ranges for wild RCP in New Zealand, as well as the first evaluation of morphometric measurements versus DNA-based sex identification in this species. There are limited published reference ranges for wild parrot species globally (McDonald et al. 2010, Le
Souef et al. 2013), and in New Zealand parrots with the exception of the kākāpō
(Low et al. 2006). The biochemistry results were limited to three main analytes commonly examined in avian patients, uric acid, bile acids, and creatine kinase.
These analytes were measured in serum as opposed to plasma; with a study in chickens finding CK results may be 21% higher (p=0.04) in serum versus plasma from the same individual, although no significant differences were reported in the same study for UA or BA (Hrubec et al. 2002).
UA is an indicator of renal function with elevations associated with dehydration or renal dysfunction (Harrison and Lightfoot 2006). As the nitrogenous end product of protein metabolism, UA reflects dietary protein content, and therefore is affected by species, age, and diet (Polo et al. 1998, Capitelli and Crosta 2013). Results can vary between captivity and the wild, and seasonally in the wild; however, there were insufficient samples in this study to infer seasonal effects. Our results for UA fell within normal ranges previously reported for wild parrots (McDonald et al. 2010), although the mean was significantly higher than that reported for the kākāpō (two tailed t-test, p<0.0001) (Low et al. 2006). Significant variation by species has also
48 been reported in captive parrots in a study of 19 species, ranging from 277 +/- 161 mmol/l in the Scarlet Macaw (Ara macao) to 853 +/- 415 umol/l in the Golden
Conure (Guaruba guarouba) (Polo et al. 1998).
Bile acids are an indicator of liver function, and may be affected by post-prandial elevations (Capitelli and Crosta 2013), as well as the method used for analysis
(enzymatic colorimetric vs radioimmunoassay) (Harrison and Lightfoot 2006). We were unable to control fasting prior to sampling, however, reference ranges in our study align with those reported elsewhere for captive parrots (ISIS 2013) and wild parrots (McDonald et al. 2010).
Our study reports absolute CK values as opposed to peak CK, with varying time from capture to sampling. Creatine Kinase results for RCP in this study were markedly higher than those reported for captive parrots (100-500 IU/L) (Capitelli and Crosta
2013), as well as limited wild parrot data (Low et al. 2006, Le Souef et al. 2013).
Creatine Kinase is an enzyme found in skeletal muscle, as well as other tissues, that elevates in cases of muscle damage (Capitelli and Crosta 2013), for example during capture and restraint. It is measured in mammals and birds to indicate likelihood of capture myopathy, a condition related to stress and overexertion during capture that may lead to muscle breakdown as a result of changes in tissue perfusion, hypoxia and hyperthermia (Williams and Thorne 1996b). The highest value reported in our study at 50-60 min post-capture (7567 IU/l) reached a level that has been associated with capture myopathy in other avian species (Businga et al. 2007).
Clinical signs of capture myopathy in birds can include incoordination, paresis or paralysis, and tachypnea (Ward et al. 2011), with more severe cases resulting in per-
49 acute, acute or chronic mortality due to metabolic derangements and/or organ failure (Williams and Thorne 1996b). Results for CK will be influenced by the capture method used, whether the bird is captive (and possibly used to handling) or wild, and species differences in tolerance of handling. However, the sensitivity of this measure as an indicator of the individual developing physiologic changes and clinical signs associated with capture myopathy is suggested to be highly species specific in birds (Ward et al. 2011), and therefore should ideally be interpreted against results from the same species. Importantly, CK levels in blood continue to elevate for 24-72 hours post capture (Bailey et al. 1997, Ward et al. 2011), therefore the time from capture to sampling will affect the result and may not indicate the peak CK value achieved. Creatine Kinase elevations of up to 15 times normal at 24 hours post capture have been noted in clinically normal Houbara bustards (Chlamydotis undulata macqueeni) (Bailey et al. 1997). It is unlikely our data reflects peak CK values in RCP, and this warrants further targeted studies in this species. No birds in our study demonstrated clinical signs of capture myopathy at release.
Most knowledge of the structure and function of avian blood cells has been derived from studies in domestic fowl (Claver and Quaglia 2009). Whilst there is reasonable homogeneity across species of these physical properties, immunologic responses and thus haemograms may vary between species (Capitelli and Crosta 2013), as well as between captive and wild counterparts of the same species, as nutritional status, age and environment also affect the haemogram (Paz Nava et al. 2001). The critical review of avian haemograms is comparatively less advanced than mammals (Claver and Quaglia 2009), and differences are reported between haemocytometers (manual
50 counts) and estimated counts from blood smears (Capitelli and Crosta 2013). Blood smears require only one drop of blood, are simple to prepare with minimal equipment in the field (Walberg 2001), can be stored long term, and provide useful information on the number and morphology of white blood cells, as well as the presence of blood parasites. However it is critical they are prepared appropriately or accuracy will be significantly affected (Walberg 2001). In comparison manual count methods, whilst more accurate, are time consuming and require a field microscope as well as training (Walberg 2001). For this study we chose the method most commonly used in the field with the species of interest, to ensure reference ranges reflected this. We recommend training by an experienced avian haematologist in the use of a slide spreader prior to conducting field studies to improve the validity and reproducibility of results.
The EWCC (using a blood smear) is an approximation of the total white cell count, and measures changes in white blood cell (WBC) numbers or morphology that are typically associated with inflammation of infectious or non-infectious origins
(Campbell and Ellis 2007). In our study, the EWCC aligned with normal ranges reported elsewhere for parrots (Polo et al. 1998). A differential count of the types of leukocytes that comprise the EWCC is important to infer the nature of any changes detected and suggest the etiology (Briscoe et al. 2010). Lymphocytes and heterophils are the most common leukocytes in the avian haemogram, however, some authors report heterophils as the dominant leukocyte in the circulation of birds generally
(Claver and Quaglia 2009), or parrots specifically (Capitelli and Crosta 2013), whilst others note the opposite with lymphocytes the principal WBC in parrots (Polo et al.
51 1998). Heterophils may respond to bacterial/viral and parasitic disease, play a role in phagocytosis, and mark an acute inflammatory response in birds (Mitchell and Johns
2008), whereas lymphocytes increase in cases of antigenic stimulation or leukaemia
(Campbell and Ellis 2007). As with other measures their relative and absolute values are likely to be species-specific, and our study found lymphocytes were more common in the haemogram of RCP, with a median relative count of 65% versus 21% for heterophils. This is important to note for accurate interpretation of results in this species, as it differs from some parrot species (Polo et al. 1998), and emphasises the risk of extrapolating from other species. Mild leukocytosis and lymphopaenia can occur with stress (Capitelli and Crosta 2013), and this may have impacted our results as all birds were captured in mist nets and handled prior to general anaesthesia.
However, these same conditions are likely to be met with future wild RCP sampling events as this is the standard capture technique.
Packed cell volume (PCV) and total protein (TP) values in the RCP were comparable to other avian species, with a median of 55% and 38 g/l respectively. Total protein in birds is generally lower than mammals, and averages 25-45 g/l (Capitelli and Crosta
2013), whilst the normal PCV range in birds is 35-55% (Mitchell and Johns 2008,
Capitelli and Crosta 2013). We chose to measure TP using a refractometer, as it is simple and transportable and therefore suited to field use. However, refractometers are less reliable than laboratory methods, and do not provide a breakdown of the types of proteins. Whilst laboratory-based plasma protein electrophoresis enhances the diagnostic utility of results, it is less commonly used by field researchers, as it relies on the facility to move samples to a laboratory in a timely fashion.
52 Culmen length and width have been described as reliable indicators of sex in RCP
(Sagar 1988), and are commonly used for this purpose in research and translocations. However no previous study has been conducted in RCP to validate sex determination based on physical characteristics with DNA-sexing methods. DNA sexing is considered a highly reliable indicator of sex (Jensen et al. 2003). Our results demonstrate that sexing by beak measurements is generally reliable, however should be interpreted with caution in the intermediary range between males and females. It is possible the five males that were sexed as females due to short beaks, had damage to their beak tips that led to incorrect sexing by morphometric measurements. In a study of wild Carnaby’s and Baudin’s black cockatoos
(Calyptorhynchus spp), 6.6% (n=12) of individuals had discrepant sex identification results between DNA sexing/endocosopic visualization of gonads and physical characteristics (Le Souef 2012). Further work in this area could include a study that incorporates endoscopic examination of gonads to confirm sex and clarify DNA versus physical characteristic results, however given the high level of agreement between non-invasive beak measurements and DNA-sexing this may not be justified.
Body weight did not vary seasonally in our study which contrasted with the findings from another study on a nearby island that found male body weights declined significantly between February and September (Sagar 1988). However female body weights were significantly higher on Hauturu-o-Toi/LBI versus Tiritiri Matangi Island, which may reflect a genuine site effect, or annual variation in resources and conditions as birds on Hauturu-o-Toi/LBI were not caught in the same year or season as birds on Tiritiri Matangi Island.
53 3.5 Conclusions
The results reported here provide conservation managers and wildlife health professionals with a set of parameters for use in determining the health status of wild RCP either in disease surveillance work, or for assessment of individual fitness for translocation purposes. Collecting reference range data for a range of physical and physiologic indices in wild birds may be influenced by multiple factors including seasonal variability, environment and resources, age, sex, physiologic status
(including stage of moult and reproductive status), operator error (measurements), and methods of sample collection, storage, transport and analysis. Interpretation of results against reference ranges must therefore take into account all these factors for both the individual being evaluated, and the ranges against which it is being compared. The notable CK elevations post-capture and handling may indicate RCP are at risk of capture myopathy with prolonged handling or restraint, and warrants further investigation. Creatine Kinase may become a useful prognostic indicator for suitability of transfer, although access to a laboratory and timely results are needed for this purpose. We demonstrated high concordance between DNA-sexing and morphometric measurements in this species, and recommend the former is performed (with or without endoscopy) if accurate sex determination is important, e.g., if assigning pairs for a breeding program.
54
CHAPTER 4:
BEAK AND FEATHER DISEASE VIRUS IN
WILD RED-CROWNED PARAKEETS
This chapter is modified from the following paper to be submitted to Archives of
Virology:
Jackson, B., Varsani, A., Holyoake, C., Jakob-Hoff, R., Robertson, I., McInnes, K.,
Empsom, R., Gray, R., Nakagawa, K., and K. Warren. 2014. Emerging infectious disease or evidence of endemicity? A multi season study of Beak and feather disease virus in wild Red-crowned Parakeets (Cyanoramphus novaezelandiae).
55 4.1 Introduction
The direct and indirect impacts of infectious disease in wildlife populations are well described and include energy costs, lowered fertility or reproductive capacity, increased susceptibility to secondary diseases (e.g. parasitism), vulnerability to predation, and mortality (Cunningham 1996). These impacts may be exacerbated by intrinsic or extrinsic factors including genetics (Altizer et al. 2003), immune status, environment (Daszak et al. 2001), and conservation management strategies that change the host-pathogen relationship or transmission rates (Thompson et al. 2008).
The ultimate cost of disease at a population scale escalates in circumstances of small or isolated populations, where Allee effects, inbreeding and other stochastic events may combine with disease to increase the likelihood of disease-mediated extinctions
(de Castro and Bolker 2005).
Due to the widespread negative impact of introduced mammalian predators on many native species, translocations and reintroductions to predator-free island and mainland sites has become a key wildlife conservation strategy in New Zealand
(Bellingham et al. 2009). These animal movements cause stress that may exacerbate existing subclinical infections (Dickens et al. 2010), as well as introducing specific disease transmission risks to source and recipient populations. Risks from multi-host pathogens extend to sympatric species, that may, in turn, alter community structures and place the whole ecosystem at risk if key species are significantly impacted (Cunningham 1996). Thus it is important to consider the indirect effects on biodiversity from pathogen spread, and their potential as invasive species in their own right (Daszak et al. 2001). Drivers of emerging infectious diseases of humans,
56 domestic animals and wildlife are primarily associated with anthropogenic environmental change, which lead to changes in host-parasite interactions that may be augmented by pathogen evolution (Daszak et al. 2001).
Beak and feather disease virus (BFDV) is a non-enveloped circular single-stranded
DNA virus from the Circoviridae family that is only known to infect parrots (Ritchie et al. 2003). The virus is host-generalist and therefore it is suspected that all species may be equally susceptible to infection (Sarker et al. 2014a). However the outcome of infection is governed by age (Doneley 2003), immune status, host species
(Schoemaker et al. 2000, Eastwood et al. 2014), as well as virus characteristics
(Kundu et al. 2012), leading to acute mortalities in some species and particularly the young (Raidal and Cross 1995). More recently, age-related host responses of apoptosis have been linked to severity of infection outcomes (Robino et al. 2014).
Chronic feather dystrophy with immune system depression is more common in older birds, while recovery or carrier status in individuals or species that mount an effective immune response can also occur (Raidal 1995). Transmission may be direct through faeces, feather dander or crop secretions, or indirect e.g. via nest site contamination. The resistant nature and long environmental persistence of BFDV
(Todd 2000), coupled with reservoir or carrier states in certain hosts that can continuously introduce virus into a system (Jackson et al. 2014b), enables this virus to operate outside of density dependent transmission models, and therefore may circumvent population thresholds for maintenance of infection in a population.
BFDV is a globally recognized conservation threat for wild parrots, with disease outbreaks reported in wild populations of the threatened Mauritius Parakeet
57 (Psittacula echo) (Kundu et al. 2012) and captive and wild populations of the critically endangered Orange-bellied Parrot (Neophema chrysogaster) (Peters et al. 2014). It has also been detected in the wild in endangered species such as Swift Parrots
(Lathamus discolor) in Australia (Sarker et al. 2014b), Cape Parrots (Poicephalus robustus) in South Africa (Regnard et al. 2014b), and Red-tailed Black Cockatoos
(Calyptorhynchus banksii) and Glossy Black Cockatoos (Calyptorhynchus lathami) also in Australia (Sarker et al. 2014a). Host-switching may drive replicative competency of
BFDV and therefore its virulence, and is likely facilitated by competition for nest- hollows in the wild (Sarker et al. 2014a). Host-switching of BFDV in captive facilities that house parrots intended for release is a substantial risk for wild populations
(Regnard et al. 2014b). Further, a high rate of recombination (Julian et al. 2013) and mutagenesis (Kundu et al. 2012, Harkins et al. 2014, Sarker et al. 2014a) of BFDV enables rapid viral evolution. This may lead to the emergence of new variants, that may have greater or altered pathogenicity within a population (Sarker et al. 2014a), leading to sudden changes in the threat level posed by this virus for wild parrot populations.
Current theory suggests the origin of BFDV was in Australia, with global anthropogenic spread via parrots in the legal and illegal trade of birds (Heard et al.
2013, Sarker et al. 2014a). This has led to speculation that the introduced Australian
Eastern Rosella (Platycercus eximius) and Sulphur-crested Cockatoo (Cacatua galerita) may have brought the virus to New Zealand (Ha et al. 2007), although there is no empirical evidence to support this. In 2008, BFDV was detected for the first time in a wild population of Red-crowned Parakeets (RCP, Cyanoramphus
58 novaezelandiae) during a translocation event (Ortiz-Catedral et al. 2009b). This finding was significant for several reasons. Some individuals demonstrated classical signs of feather loss that usually progress to chronic disease and mortality; the closely related variants and high prevalence identified supported a recent introduction or outbreak (Ortiz-Catedral et al. 2010, Massaro et al. 2012), and the location was a source population for translocations to other regions in New Zealand.
BFDV is considered endemic in free-living exotic parrots on the North Island of New
Zealand (Ha et al. 2007). Surveillance for BFDV infection across New Zealand following the presumed outbreak reported by Luis et al. (2009) provided further evidence of a relatively high endemic prevalence in Eastern Rosellas, as well as ongoing detections in RCP on Hauturu-o-Toi/Little Barrier Island (LBI) (Massaro et al.
2012). However sampling failed to detect BFDV in native kākā (Nestor meridionalis), kea (Nestor notabilis), or the critically endangered kākāpō (Strigops habroptilus) and
Malherbe’s Parakeet (OFP, Cyanoramphus novaezelandiae) (Massaro et al. 2012).
Interestingly, Massaro et al. (2012) reported a second introduction of BFDV in
Yellow-crowned Parakeets (YCP) in the remote Fiordland region of the South Island.
Given the lack of exotic parrots in this region specifically and the South Island generally, this raises the possibility of co-evolved or endemic BFDV variants in New
Zealand, as the chances of introduction by exotic parrots in this region appear unlikely.
Whilst there are multiple studies describing BFDV detection in New Zealand (Ha et al.
2007, Ha et al. 2009, Ortiz-Catedral et al. 2010, Massaro et al. 2012), most report point or period prevalences that may extend over long time frames. These limit our
59 ability to describe disease trends temporally or spatially, thereby to infer factors that underpin BFDV emergence, or population scale risks. Critically, we need to determine whether BFDV is an emerging infectious disease of native parrots in New
Zealand or whether it has been present historically such that species may have co- evolved with or adapted to this pathogen.
This study aimed to examine the seasonal trends of BFDV infection in a wild population of RCP; the agreement between blood and feather samples for BFDV detection; the evidence of flock immunity or exposure through testing for antibodies
(haemagglutination inhibition test; HI); and the phylogenetic relationships to circulating BFDV in Eastern Rosellas in New Zealand. The results contribute to our understanding of the conservation significance of BFDV in RCP specifically, and, more broadly, for the Cyanoramphus genus regionally and parrots globally. This updated knowledge is particularly important for local translocation and reintroduction programs, which are currently restricted by a lack of empirical evidence of the impact of BFDV, and the likely source of virus into New Zealand. This study also presents the first published seroprevalence data for BFDV in a New Zealand parrot species, to further examine disease dynamics and flock immunity in this species.
Through the course of this study, opportunistic or passive surveillance samples collected from three other sites on the North Island were tested, and these are included for molecular epidemiological analyses and interpretation.
60 4.2 Methods
Study sites
Tiritiri Matangi Island is a 220 ha open scientific reserve with ongoing ecological restoration since it was established in 1980 in the Hauraki Gulf of New Zealand
(Figure 4.1). The island is jointly managed by the Department of Conservation (DOC) and the Supporters of Tiritiri Matangi (SOTM) community group (Galbraith and
Cooper 2013). A population of RCP was established from captive bred birds released at this site between 1974-1976 (Ortiz-Catedral and Brunton 2010b).
Hauturu-o-Toi/Little Barrier Island (LBI) is a 2817 ha ecologically intact and closed scientific reserve established in 1985 and managed by DOC, also in the Hauraki Gulf
(Figure 4.1). There is a resident remnant population of RCP on the island.
Capture and sampling methods
RCP were captured on Tiritiri Matangi Island using mist nets (Avinet, USA) during four cross sectional sampling sessions in April and September of 2011 and 2012. RCP were captured with the same methods on Hauturu-o-Toi/LBI in June 2013. All birds were taken to a field station in single use calico bags (Prospectors Earth Sciences,
Cairns, Australia), and anaesthetized with isoflurane and oxygen from a portable field anaesthetic machine (AAS, Sydney, Australia). Approximately 0.1ml of blood was collected from the medial metatarsal vein, and 2-4 growing feathers from the breast region. Blood was placed on Whatmanns No.3 filter paper and two blood smears were made on glass slides using a customised spreader to improve smear quality. Feathers were preserved in 70% ethanol. A full physical examination was
61 conducted including auscultation of the heart and airsacs/lungs, body condition was scored (poor, average, good, very good), standard morphometric measurements
(culmen width/length, flattened wing chord and tarsometatarsus) were obtained, and standardised photos of the head and body were taken. All birds were banded with metal and colour leg bands (DOC, Wellington, NZ), and recovered in small plastic bird boxes prior to release.
Given the persistence and stability of BFDV, specific precautions were taken to prevent cross-contamination of samples, sample pots, materials, and also transmission of any pathogens between individuals. All instruments, anaesthetic equipment, calico bags, and carry boxes were cleaned then soaked in Trigene®
(Ethical agents, Auckland, New Zealand) at 1:100 dilution and left in contact for a minimum of 5 minutes between birds (or overnight for calico bags). The veterinarian and nurse wore latex gloves, and changed these between each bird. A designated
‘clean’ field assistant was used to record data, and label and present sample pots or bags to the veterinarian.
Passive surveillance samples
Between July 2010 and September 2012, 2-5 body contour feather samples per individual were collected opportunistically from RCP handled during field activities at
Kapiti Island, ZEALANDIA-Karori sanctuary (hereafter ZEALANDIA), and Matiu-Somes
Island (Figure 4.1). Feathers were stored in a paper envelope at room temperature.
One feather sample from a RCP found dead with a history of feather loss was obtained from Shakespear Regional Park in Auckland, New Zealand (Figure 4.1).
62
Figure 4.1: Map of the North Island of New Zealand showing sites where samples for
Beak and feather disease virus screening were obtained.
Haemagglutination Inhibition (HI) assay
Haemagglutination Inhibition assays (HI) were outsourced and carried out by Charles
Sturt University (Australia) on stored frozen blood on filter paper using previously reported methods (Riddoch et al. 1996). Based on previous studies using HI (Raidal et al. 1993a, Khalesi et al. 2005 ), a titre of 1:20 was set as the minimum threshold for a positive result.
Molecular testing
DNA was extracted from feathers and blood on filter paper with the iGenomic blood
DNA extraction kit (Intron Biotechnology, South Korea) according to the
63 manufacturer’s instructions. PCR testing for BFDV was performed with 4µl of the extracted DNA, using the KAPA blood PCR kit Mix B (KAPA Biosystems, USA) and primers that target a ~605bp region of the replication associated protein (Rep) gene
(Julian et al. 2012). Recovery of full genomes from BFDV positive samples also followed previously described methods (Varsani et al. 2011 ), with 1µl of extracted
DNA enriched using TempliPhiTM (GE Healthcare, USA), and used as template for PCR amplification with back-to-back primers (BFDV-AV-F 5’-CYT ACY CTK GGC ATT GTG
GC-3’, BFDV-AV- R 5’-TAT HAC RTC BCC YTC YTT ACT GCA-3’) and Kapa HiFi HotStart
DNA polymerase (Kapa Biosystems, USA). These amplified full genomes were cloned into a pJET1.2 vector (Thermofisher, USA) and the recombinant plasmid sequenced by primer walking at Macrogen Inc. (Korea).
Sequences were assembled using DNAMAN (version 7, Lynnon Biosoft, Canada), and the full genomes aligned with those available in GenBank using MUSCLE (Edgar
2004). A maximum likelihood phylogenetic tree was inferred using PHYML (Guindon et al. 2010) using a GTR+I+G4 nucleotide substitution model, selected as the best substitution model by jModelTest (Posada 2008) and with approximate likelihood ratio test (aLRT) branch support. Branches with <84% aLRT branch support were collapsed using MESQUITE version 2.75 (Maddison and Maddison 2011). Percentage pairwise identities were calculated using SDT v1.0 (Muhire et al. 2013).
Epidemiological design and analyses
Cross-sectional studies on Tiritiri Matangi Island and Hauturu-o-Toi/LBI aimed to capture up to 65 individuals to provide a target confidence of detection of BFDV down to 5% prevalence with 95% confidence, assuming test sensitivity of 0.9 and
64 specificity of 1 (positives were only reported if a full genome sequence could be obtained). A minimum capture of 33 individuals was set for 95% confidence of detecting BFDV if it was present at or above 10% prevalence, with the same test characteristics. Samples sizes were calculated in Epitools (Sergeant 2015).
Descriptive statistics for prevalence data were calculated in Microsoft Excel. All measures of significance were conducted in R-studio version 0.98.501.
4.3 Results
BFDV PCR prevalence and HI-based seroprevalence
Beak and feather disease virus PCR prevalence data by season and location is summarised in Table 4.1, including the target study sites of Tiritiri Matangi Island and
Hauturu-o-Toi/LBI, as well as the opportunistic sampling series from Kapiti Island,
ZEALANDIA sanctuary, and Matiu-Somes Island. Prevalence data is not reported for the Shakespear Regional Park sample since only a single sample was collected. The total prevalence for Tiritiri Matangi Island across all sampling sessions was 1.09%
(95%CI: 0.1-3.9%), however positive birds were only detected in the first sampling season. HI seroprevalence results are also summarized in Table 4.1, with Table 4.2 providing further details on the titres for HI positive individuals.
65 Table 4.1: PCR and Haemagglutinaion Inhibition (HI) titre prevalence results for BFDV testing of Red-crowned Parakeets at various locations on the North Island of New
Zealand during the study period. PCR positive results are from blood and/or feather samples.
Location Season BFDV HI BFDV PCR n
Tiritiri Matangi Island April 2011 26% 3.8% 53
(15.3-40.3%) (0.5-13.0%)
Sept 2011 NA 0% 34
(0-10.3%)
March 2012 NA 0% 38
(0-9.3%)
Sept 2012 NA 0% 59
(0-6.1%)
Hauturu-o-Toi/LBI June 2013 14% 4.4% 45
(5.3-27.9%) (0.5%-15.1%)
Kapiti Island July 2010 – NA 3.4% 146 Sept 2012 (1.1-7.8%)
ZEALANDIA Jan 2011 – NA 1.6% 64 May 2012 (0-8.4%)
Matiu-Somes Island June 2012 NA 0% 28
(0-12.3%)
66 Table 4.2: Individual haemagglutination inhibition (HI) titre results from testing of
Red-crowned Parakeets on Hauturu-o-Toi/LBI and Tiritiri Matangi Island
HI test results
Location n <1:20 1:20 1:40 1:80 1:160 1:640 1:1280
Hauturu-o-Toi/LBI 43 37 0 0 1 3 1 1
Tiritiri Matangi Island 53 39 13 1 0 0 0 0
BFDV positive birds
Only blood samples yielded BFDV positive results by PCR, with all feathers testing negative. Table 4.3 summarises data on BFDV positive birds from the cross-sectional studies on Tiritiri Matangi Island and Hauturu-o-Toi/LBI, including results of HI testing. Although three males tested positive for BFDV versus a single female, there was not a significant difference associated with gender (p=0.342).
Table 4.3: BFDV positive Red-crowned Parakeets from Tiritiri Matangi Island and
Hauturu-o-Toi/Little Barrier Island, including sex, location, blood and feather PCR results, and Haemagglutination Inhibition (HI) results. An HI result of <1:20 was defined as a negative.
Location Sex Health PCR PCR HI test status blood feather
Bird 1 Tiritiri Matangi Island M Normal + - <1:20
Bird 2 Tiritiri Matangi Island M Normal + - <1:20
Bird 3 Hauturu-o-Toi/LBI M Normal + - <1:20
Bird 4 Hauturu-o-Toi/LBI F Normal + - 1:160
67 Table 4.4 (end of chapter) details BFDV prevalence data for wild parrots in New
Zealand, including previous reports and those from this study. Where possible, prevalence estimates have been stratified by sampling sessions or years, to demonstrate inter-annual variation in some species such as RCP.
BFDV molecular analyses
All strains identified fall within the monophyletic clade of the BFDV-A strain. A maximum likelihood phylogenetic tree of the BFDV-A strain is presented in Figure
4.2, and full genome sequences have been deposited in GenBank (KM452734 -
KM452744). The BFDV genomes isolated in this study share >95% pairwise identity with other BFDV genomes reported from the North Island of New Zealand. It is important to note that BFDV-A strains have only been identified in the North Island of New Zealand infecting RCP and Eastern Rosella. There are two distinct and well- supported clades of BFDV-A variants in RCP on Hauturu-o-Toi/LBI, one for isolates sampled in other studies between 2008-2010 (n=16) (Massaro et al. 2012), and the second for samples from this study in 2013 (n=2). The isolates of BFDV from RCP sampled on Kapiti Island (n=5) and ZEALANDIA (n=1) form a distinct well-supported clade. The two isolates of BFDV from RCP sampled on Tiritiri Matangi Island share a most common ancestor to two Eastern Rosella derived isolates sampled in 2011
(Jackson et al. 2014a). It is also important to note that the genome of a single isolate of BFDV-A from Shakespear Regional Park in Auckland was found to be well nested within BFDV-A variants recovered from Eastern Rosella sampled in Auckland
(Massaro et al. 2012, Jackson et al. 2014a). We did not find any additional evidence
68 of recombination within the BFDV-A isolates other than that previously described in
Julian et al. (2012).
69 GU936290 (2008) ) 100
Eastern rosella % (
GU936294 (2008)
y t
Red-crowned parakeets i 99
GU936288 (2008) t n
GU936296 (2008) e 99 d
Hauturu-o-Toi/LBI i
GU936295 (2008) e s
i 98
Tiritiri Matangi Island GU936293 (2008)
w r
V i
Kapiti Island GU936297 (2008) a 98 100 93 GQ396655 (2008) P Zealandia Sanctuary 97 GU936291 (2008) Auckland GU936292 (2008) 96 FJ519618 (2010) 96 GQ396653 (2008) 92 GQ396656 (2008) 95 GQ396652 (2008) 95 GU936289 (2008) 85 GQ396654 (2008) 94 JQ782198 (2010) 100 JQ782199 (2010) KF467251 (2012) 94 KF467252 (2012) 99 KM452734 (2013) 100 FJ519619 (2010) JQ782196 (2009) 99 99 KM452735 (2013) 100 KM452736 (2013) 97 GU936287 (2008) 99 KF467253 (2012) 96 KF467254 (2012) 98 KM452737 (2011) KM452738 (2011) 100 KM452739 (2012) 85 KM452740 (2012) KM452741 (2012) KM452742 (2010) 100 KM452743 (2010) KM452744 (2009) JQ782197 (2009)
92 JQ782200 (2009)
) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) )
) ) ) ) ) )
) ) ) ) )
8 8 8 8 8 8 8 8 8 8 0 8 8 8 8 8 0 0 2 2 0 9 8 2 2 9 9
3 0
3 3 1 1 2 2 2 0 9
0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0
1 1
1 1 1 1 1 1 1 1 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (
2 2 2 2 2 2 2 2 2 2 2
0.005 nucleotide substitutions per site
( ( ( ( ( (
( ( ( ( (
0 4 8 6 5 3 7 5 1 2 8 3 6 2 9 4 8 9 9 6 7 7 0
1 2 3 4
9 0
9 9 8 9 9 9 9 5 9 9 1 5 5 5 8 5 9 9 1 9 8
4 5 6 8 9 0 1 2 4
7 3
5 5 5 5
1 2
2 2 2 2 2 2 2 6 2 2 6 6 6 6 2 6 1 1 6 1 2
2 2 2 2
3
3 3 3 3 3 4 4 4 4 4
2 2
6 6 6 6 6 6 6 6 6 6 9 6 6 6 6 6 2 2 9 2 6
7 7 7 7
7 7 7 7 7 7 7 7 7 7 7
3 3 3 3 9 3 8 8 8
3 3 3 3 9 3 1 9 9 9 8 1 8 3
6 6 6 6
9 9 9 9 9 9 9 3 9 9 5 3 3 3 9 3 7 7 5 7 9 7 7
2 2 2 2 2 2 2 2 2 2 2
4 4 4 4
J J
U U U U U U U U U U U
5 5 5 5 5 5 5 5 5 5 5
Q Q Q Q Q Q Q Q Q Q
F F F F
F F
J J J J J
4 4 4 4 4 4 4
4 4 4 4
G G G G G G G G G G G
G G G G G K K K K
M M M M M M M
M M M M
K K K K K K K K K K K
Figure 4.2: Maximum likelihood phylogenetic tree of the full genomes of BFDV isolates on the North Island of New Zealand (BFDV-A strain),
showing the relationship between Eastern Rosella (purple line) and Red-crowned Parakeet (green line) isolates. Pairwise identities (%) are
presented on the right of the tree, with all isolates sharing ≥95% pairwise identity. BFDV isolates from this study are in bold font and year of
sampling is provided in brackets after the GenBank accession numbers.
70 4.4 Discussion
Phylogenetic analysis of the BFDV genomes from this study show ongoing viral flow between RCP and Eastern Rosellas on the North Island of New Zealand, with separate but closely related variants circulating in Kapiti/ZEALANDIA in the South of the North Island (Figure 4.2). The two distinct clades of BFDV-A variants in RCP on
Hauturu-o-Toi/LBI may indicate two separate introductions to the population.
However, the role of RCP or Eastern Rosellas as the source of BFDV in New Zealand cannot be resolved. Data from this and other studies suggest BFDV persists at a higher PCR prevalence in exotic parrots (range 14.8-28%) (Ha et al. 2007, Massaro et al. 2012) than RCP who generally show a low level (range 1.1-4.4%) or undetectable
PCR prevalence. Two notable spikes in prevalence have been detected in RCP in 2008
(19.7%) and YCP in 2012 (26.7%) (Massaro et al. 2012), although a lack of consistent monitoring in these populations makes it difficult to be sure they are truly above baseline. These spikes may have been associated with environmental or population changes that increase contact rates and the proportion of susceptible individuals in a population. Although we did not detect BFDV in 3 of 4 cross-sectional studies on
Tiritiri Matangi Island, our sample sizes were insufficient to provide statistically significant confidence of absence of BFDV infection in that population during that time period. Similarly, the sample size (n=28) from Matiu-Somes Island means we cannot be statistically confident this population is free from a low prevalence of
BFDV infection. This highlights both the importance of appropriate sample sizes to prevent falsely reporting an infection-free population, and the difficulty of detecting
71 low prevalence diseases in wildlife populations due to the large sample sizes required.
Results from HI testing are the first to demonstrate that wild RCP have the capacity to develop an immune response to BFDV. A high HI titre in an individual is a negative predictor for chronic BFDV infection, and positively correlated with immune status
(Raidal et al. 1993b, Khalesi et al. 2005 ). In this study, only one of four BFDV PCR positive individuals had a positive HI result, with the other birds either caught at an early stage of infection prior to antibody development, or failing to develop antibodies with subsequent disease likely. The study also found a relatively low seroprevalence (14 - 26%), in conjunction with low PCR prevalence, for BFDV in both populations studied. In endemically infected flocks in Australia, low PCR prevalence of BFDV infection in populations has been associated with high seroprevalence (41-
94%) (Raidal et al. 1993a), indicating high exposure and the development of flock immunity. Our results may relate to a failure of transmission within RCP flocks (due to flock dynamics and contact rates), and therefore low overall BFDV exposure within the population. They may also indicate that RCP mount a transient or undetectable immune response to BFDV when clearing the virus. Overall, the data suggests RCP may be a dead-end or spill over host, with ongoing exposure to BFDV reliant on either a reservoir species (e.g., Eastern Rosella) or repeat introductions from environmental sources (e.g., infected nest sites). Conversely, low PCR prevalence and low HI seroprevalence could be interpreted as a population highly susceptible to disease with individuals removed rapidly from the population for
72 sampling due to high mortality. To date there has been no evidence of the latter hypothesis in closely monitored populations such as that on Tiritiri Matangi Island.
A number of epidemiological factors govern BFDV dynamics, with reported variation in species and individual host susceptibility (Doneley 2003) giving rise to the hypothesis that some New Zealand parrots may in fact be generally resistant to fatal forms of the disease. Although clinical disease including loss of wing and tail feathers has been reported in RCP and YCP of varying age classes in wild populations (Ortiz-
Catedral et al. 2009b, Massaro et al. 2012), evidence of widespread clinical disease in parakeet populations has not been observed to date, even in well-monitored or observed populations. Life history traits play a role in disease dynamics and population recovery; e.g. epidemics may occur in highly fecund species due to the sudden increase in susceptible individuals and population density following a breeding event (Lloyd-Smith et al. 2005). However high reproductive potential may also ensure a population recovers more rapidly from a mortality event. RCP, and
Cyanoramphus sp. in general, respond to masting events of local food sources (e.g. flax or beech) with markedly increased breeding (Elliot et al. 1996) that may play a role in sporadic disease emergence above endemic or baseline prevalence, and this hypothesis warrants further investigation. The population impacts of BFDV infection in a species with low reproductive capacity, such as the kākāpō, are currently unknown and may be significantly higher. Thus it is important to interpret the potential impacts of this disease at the population level with the ecology and life history traits of the species of interest (Armstrong and Ewen 2013).
Reservoirs of disease will counteract population threshold limitations (Daszak et al.
73 2001), as transmission via reservoir individuals or species, or contaminated sites, act outside of density dependence (Storm et al. 2013). BFDV is highly resistant to disinfection, persists for years in the environment (Todd 2000), with some species appearing more likely to enter carrier states that enable viral excretion into the environment for extended periods. Thus there are biotic (e.g. carrier host or species) and abiotic (eg nest site) reservoirs (de Castro and Bolker 2005), that likely enable persistence of BFDV beyond population threshold limits for direct transmission.
Management factors such as nest boxes and feeding stations can artificially increase contact rates, host density and interspecies transmission by encouraging congregation of birds (Kirkwood 1998). On Tiritiri Matangi Island, supplementary water sources are frequented by RCP, and artificial nest boxes have been supplied with different species using the same nest box in some cases. In this study prevalence of BFDV was not higher on Tiritiri Matangi Island, where high population density and management factors might increase the likelihood of virus transmission, as compared to Hauturu-o-Toi/LBI where available habitat supports a more dispersed population. However, RCP on Hauturu-o-Toi/LBI are noted to aggregate at edge habitat that may create locally increased population densities based on resource availability.
The question as to the origins of BFDV in New Zealand remains unanswered. There are four plausible pathways by which BFDV may have become established in the wild in New Zealand, and it is possible all four have contributed to BFDV spread, and will continue to do so in the future. These are: (1) co-evolution with the native parrots of
New Zealand, (2) deliberately introduced infected exotic parrots that established in
74 the wild (Eastern Rosella and Sulphur-crested Cockatoo), (3) infected captive-bred native parrots released to the wild during reintroductions, and (4) accidental or deliberate releases of infected captive parrots (native or exotic). On Tiritiri Matangi
Island, 80 captive-bred RCP were released as founder birds between 1974 and 1976
(Ortiz-Catedral and Brunton 2010b), prior to the availability of PCR testing for BFDV, therefore their disease status at the time of movement is unknown. The species, geographic and temporal distribution of BFDV detections, combined with the molecular evidence, does not support a single source of virus to the wild in New
Zealand. It appears likely that the historical movement of parrots to, and within, New
Zealand has contributed to the circulating strains of BFDV in the wild, at least in the
North Island.
Disease surveillance methods will influence detection probabilities for BFDV, and thus interpretation of its conservation significance. Study designs should incorporate an appropriate temporal framework to detect cyclical trends in prevalence data. Our data supports a long term or secular trend in disease prevalence, biased to the host, with disease prevalence either stable or declining over time (Thrusfield 2007). As other studies have demonstrated (Kundu et al. 2012), long term datasets are useful to detect if an epidemic occurs, and determine what molecular, host or environment factors may precede the event. Sample type also affects detection probabilities.
Although feather has been reported to be more sensitive than blood for BFDV PCR testing in one study (Hess et al. 2004), our study and others have found blood to be more sensitive (Khalesi et al. 2005 ), and yet others have reported varying results between the two (Ritchie et al. 2003, Ha et al. 2009). We recommend testing both
75 feather and blood by PCR to improve the sensitivity of BFDV detection, ideally combined with complementary tests such as HI and haemagglutination (HA) to detect immunity and virus excretion respectively (Khalesi et al. 2005 ).
Given this study and others have found that BFDV infection in wild RCP and Eastern
Rosella on the North Island is widespread, the ability to prevent viral range expansion through movement constraints or bans on translocations becomes questionable. However, there are fitness costs of BFDV infection in an individual, and it is difficult to determine at the point of transfer whether a BFDV positive individual will go on to develop clinical disease. Our study demonstrates that clinically normal or apparently healthy animals should not be assumed to be infection-free for translocations (Cunningham 1996), as all BFDV positive birds were normal on physical examination. Moving a known infected bird means progeny of that bird, and those that travel with it, are likely to be exposed to BFDV which may impact on their survival or fecundity resulting in a reduction in the size of the effective founder population. Also, the risks of combining different BFDV strains and the potential for recombination, mutations and increased virulence are well described (Kundu et al.
2012, Julian et al. 2013, Sarker et al. 2014c). Theoretically a novel strain introduced into a population could lead to an outbreak event in a previously endemically infected and stable population. We argue against moving birds with BFDV for these reasons, and robust testing prior to bird movements is therefore recommended despite widespread BFDV infection on the mainland.
4.5 Conclusions
Can or should we manage BFDV in New Zealand parrots? The threat posed by BFDV
76 must be examined within the context of the species and system at hand, with small, critically endangered, naïve or isolated populations most at risk from disease- mediated extinctions (Peters et al. 2014). Our study has provided evidence of ongoing viral flow between RCP and Eastern Rosella on the North Island of New
Zealand, suggesting naïve populations are unlikely to be common in this region. The future of parrot conservation in New Zealand will need to incorporate BFDV as an ongoing risk, and should be vigilant for outbreak events, however there is now sufficient evidence to suggest in parakeets the threat has been downgraded since the initial detection. Species that are critically endangered with low reproductive potential (e.g. kākāpo) should still be considered at high risk in the absence of proof that they are not susceptible, and biosecurity to prevent the establishment of reservoir species (e.g. Eastern Rosella) in their habitat should remain a priority.
Given recent studies showing correlations between increased viral loads and clinical expression of disease (Regnard et al. 2014a), and decreased viral loads and potential host resistance in hybrid species (Eastwood et al. 2014), future BFDV research in
New Zealand should compare viral loads, physical examination findings and prevalences between infected species and populations in New Zealand.
77 Table 4.4: Selected reported BFDV prevalence data for wild exotic and native parrots in New Zealand
Species Common name Location(s) sampled Year(s) of PCR prevalence (95%CI) Sample size Reference sampling Cacatua galerita Sulphur-crested Cockatoo North and South Island 2001-2004 28% (22.6%-33.9%) 255 Ha et al. 2007 Platycercus eximius Eastern Rosella North and South Island 2004-2006 14.8% (9.7%-21.2%) 162 Ha et al. 2007 Platycercus eximius Eastern Rosella Auckland region 2009-2010 23.3% (9.9-42.3%) 30 Masssaro et al. 2012 Nestor meridionalis kākā North and South Island 2003-2005 0% (0-4.6%) 79 Ha et al. 2009 Nestor meridionalis kākā Great Barrier and Little Barrier Island 2008-2010 0% (0-33.6%) 9 Masssaro et al. 2012 Nestor meridionalis kākā Eglinton 2004-2011 0% (0-17.6%) 19 Masssaro et al. 2012 Nestor notabilis kea South Island 2005-2010 0% (0-3.8%) 95 Masssaro et al. 2012 Strigops habroptilus kākāpo Codfish Island 2003-2005 0% (0-8.2%) 43 Ha et al. 2009 Strigops habroptilus kākāpo Codfish Island 2009-2010 0% (0-4.2%) 87 Masssaro et al. 2012 Cyanoramphus auriceps Yellow-crowned Parakeet Hauturu-o-Toi/LBI 2010 0% (0-33.6%) 9 Masssaro et al. 2012 Cyanoramphus auriceps Yellow-crowned Parakeet Eglinton 2011-2012 26.7% (12.3-33.6%) 30 Masssaro et al. 2012 Cyanoramphus auriceps Yellow-crowned Parakeet Marlborough Sound 2003-2005 0% (0-13.7%) 25 Ha et al. 2009 Cyanoramphus novaezelandiae Red-crowned Parakeet North and South Island 2003-2005 0% (0-16.8%) 20 Ha et al. 2009 Cyanoramphus novaezelandiae Red-crowned Parakeet Hauturu-o-Toi/LBI 2008 19.7% (11.2-30.9%) 71 Masssaro et al. 2012 Cyanoramphus novaezelandiae Red-crowned Parakeet Hauturu-o-Toi/LBI 2009 2.2% (0.1-11.5%) 46 Masssaro et al. 2012 Cyanoramphus novaezelandiae Red-crowned Parakeet Hauturu-o-Toi/LBI 2010 2.8% (0.1-14.5%) 36 Masssaro et al. 2012 Cyanoramphus novaezelandiae Red-crowned Parakeet Hauturu-o-Toi/LBI 2013 4.4% (0.5-15.1%) 45 This study Cyanoramphus novaezelandiae Red-crowned Parakeet Codfish Island 2008-2010 0% (0-10.9%) 32 Masssaro et al. 2012 Cyanoramphus novaezelandiae Red-crowned Parakeet Tiritiri Matangi Island 2004-2010 0% (0-4.4%) 82 Masssaro et al. 2012 Cyanoramphus novaezelandiae Red-crowned Parakeet Tiritiri Matangi Island 2011-2012 1.1% (0.1-3.9%) 184 This study Cyanoramphus novaezelandiae Red-crowned Parakeet Kapiti Island 2010 –2012 3.4% (1.1-7.8%) 146 This study Cyanoramphus novaezelandiae Red-crowned Parakeet Matiu-Somes Island 2012 0% (0-12.3%) 64 This study Cyanoramphus novaezelandiae Red-crowned Parakeet ZEALANDIA 2011 – 2012 1.6% (0-8.4%) 28 This study
78 CHAPTER 5:
HISTOPATHOLOGY OF
KNEMIDOKOPTINID MANGE IN WILD
RED-CROWNED PARAKEETS 5
10
This chapter is modified from the following paper in review with the Journal of
Wildlife Diseases:
Jackson, B., Heath, A., Harvey, C., Holyoake, C., Jakob-Hoff, R., Varsani, A.,
Robertson, I., and K. Warren. 2014. First report of knemidokoptinid (Epidermoptidae:
Knemidokoptinae) mite infestation in wild red-crowned parakeets (Cyanoramphus 15 novaezelandiae); correlations between macroscopic and microscopic findings.
79 5.1 Introduction
Mites are the primary cause of parasitic skin and feather disease in birds (Harrison and Lightfoot 2006, Fletcher 2008), and have evolved to adapt to different environments, conditions and hosts. The most prevalent group of mites in domestic 20 birds are in the family Epidermoptidae, subfamily Knemidokoptinae (Harrison and
Lightfoot 2006). These epidermoptid or skin mites cause mange across a broad avian assemblage including passerines, galliformes and psittacine (Wade 2006). Their morphology and biology resemble the mammalian parasites, Sarcoptidae, with all life stages being present on the host and requiring direct or close contact for 25 transmission (Müllen and Durden 2009).
There are six genera and 17 species in the Knemidokoptinae (Fain and Elsen 1967,
Fain 1974, Fain and Lukoschus 1979, Mironov et al. 2005). Historically, the literature has focused on those of significance to poultry and caged birds, including ‘scaly face’
(Knemidocoptes pilae, Lavoipierre and Griffiths, 1951) in budgerigars (Melopsittacus 30 undulatus) (Yunker and Ishak 1957), de-pluming itch (Neocnemidocoptes gallinae and Picicnemidocoptes laevis Pence, 1972) in poultry and pigeons respectively (Fain and Elsen 1967) and scaly leg (K. mutans, Robin and Lanquetin, 1859) in poultry
(Morishita et al. 2005). Knemidokoptinid mange has been considered uncommon in wild birds (Mainka et al. 1994), however in the past 15 years there has been an 35 increasing number of reports documenting these parasites in a range of wild birds.
Passeriformes are well represented (Mainka et al. 1994, Ladds 2009, Dabert et al.
2011, Dabert et al. 2013); including lyre birds (Menura sp), currawongs (Strepera graculina), and forest ravens (Corvus tasmanicus) in Australia (Mason and Fain 1988,
80 Jaensch et al. 2003, Holz et al. 2005), Amakihi (Hemignathus virens) in Hawaii 40
(Gaudioso et al. 2009), and stitchbirds (Notiomystis cincta) in New Zealand (Low et al. 2007).
Clinical signs of knemidokoptinid mange vary based on the host and parasite species involved (see Table 5.1). Knemidokoptes pilae infestation is most common in budgerigars, causing hyperkeratotic encrustations on the beak, cere and legs (Koski 45
2002) that may affect perching or cause respiratory compromise in severe cases
(Ladds 2009). Severe hyperkeratosis leading to loss of whole digits and feet have been reported in American robins (Turdus migratorius) during a mange epizootic due to K. jamaicensis (Turk, 1950), with associated perching difficulties and reluctance to feed (Pence et al. 1999). Although site predilection varies between species, mites are 50 generally restricted to featherless areas such as the beak, legs and feet, and characterised by hyperkeratosis of varying severity (Wade 2006). Proliferative papillary-like lesions are also reported (Kirmse 1966, Schulz et al. 1989), and may be confused with other disease entities such as avian pox (Kirmse 1966), and papillomavirus infection (Literák et al. 2003). Notable exceptions to this standard 55 clinical presentation include de-pluming itch of fowl (N. gallinae) and pigeons (P. laevis), where lesions are commonly on the body and the main clinical sign is loss of feathers rather than hyperkeratosis (Fain and Elsen 1967).
Knemidokoptinid mange lesions are mostly confined to the stratum corneum and are characterized by mild to severe compact orthokeratotic hyperkeratosis and 60 acanthosis (Fletcher 2008). Often numerous tunnels are formed within the crusts and filled with mites (Blackmore 1963, Mason and Fain 1988), lending a
81 honeycombed appearance that may be evident macroscopically. Dermal changes are less common, although may include a perivascular mononuclear infiltrate
(Mason and Fain 1988, Ladds 2009), or occasionally a heterophilic infiltrate (Fletcher 65
2008). Subcorneal pustules and mites burrowing beneath the stratum corneum have also been reported (Low et al. 2007).
Here we present findings from an investigation of feather loss in two island populations of Red-crowned Parakeet (RCP, Cyanoramphus novaezelandiae) where pathological changes were linked to a knemidokoptinid mite. We aimed to describe 70 normal RCP skin, the pathology associated with feather loss, characterise the stages of mange in this species, and describe key features that enabled identification and differentiation of the mite species from similar members of the Knemidokoptinae.
Table 5.1: Avian families commonly infested by the Knemidokoptinae, with species of mites found and area of the body affected during clinical disease. Translated and 75 adapted to reflect current nomenclature from Fain and Elsen (1967).
Host family Knemidokoptinae species reported Areas affected Galliformes Knemidokoptes mutans Legs Neocnemidocoptes gallinae* Body
Columbiformes Picicnemidocoptes laevis** Body
Psittaciformes Knemidokoptes pilae Beak, legs and rarely skin of body Procnemidocoptes janssensi Body
Passeriformes Knemidokoptes fossor Base of beak Knemidokoptes jamaicensis Legs Knemidokoptes intermedius Legs
* Formerly reported as Neocnemidocoptes laevis gallinae
** Formerly reported as Neocnemidocoptes laevis laevis
82 5.2 Materials and Methods 80
Location and sampling design
Sampling took place in March and September 2012 on Tiritiri Matangi Island (S36˚
36' 2", E174˚ 53' 24"), and June 2013 on Hauturu-o-Toi/Little Barrier Island (LBI, S36°
11' 32, E175° 4' 29) in the Hauraki Gulf of New Zealand. RCP were captured using standard mist nets (9m x 2.6m, 30mm mesh, Avinet, USA), placed in single-use 85 cotton bags (Prospectors Earth Sciences, Australia) and taken to a field processing station. Birds were anesthetized with isoflurane at an initial flow rate of 5% in 1L oxygen, reducing to 1-2% isoflurane in 1L oxygen after approximately 30 seconds. A zero dead-space circuit was used to minimize anesthetic risk (Advanced Anesthesia
Specialists, Australia). Samples were collected for other studies, including blood, 90 feathers and feces. A skin biopsy (3-5mm) was taken from the caudal aspect of the base of the head, using fine sharp scissors and forceps, and placed in 10% neutral buffered formalin. This site was chosen as the head was a common place for feather loss, and could be easily standardized across birds whilst being in a better location for healing than other areas of the head as it was under less tension and less 95 exposed. In a subset of severely affected individuals from Tiritiri Matangi Island
(n=7), a section of skin was stored frozen in 70% ethanol for later extraction and identification of any skin parasites. Biopsy sites were closed with a single suture of 4-
0 absorbable suture material (Visorb, SVS, New Zealand), and antibacterial ointment applied. All birds were banded using metal and color bands (Department of 100
Conservation, NZ). No birds were resampled within a sampling session. A standard set of morphometric measurements were collected, and photos of the head
83 (right/left lateral and front), and body (ventral and dorsal aspect with wings extended) were taken. Birds were recovered in plastic bird-boxes on a fleece pouch, and released by the attending veterinarian or nurse. All birds were provided with 105 heat pads (wheat bags) during and after anesthesia when appropriate, and were given 0.9% normal saline (NaCl) subcutaneously at a rate of 1% of body weight.
Feather loss
A grading system was used to categorise feather loss by severity, with the following descriptors; 0 = no feather loss, 1 = patchy or focal feather loss with no coalescing 110 areas, 2 = large areas of feather loss <50% of head, 3 = >50% feather loss on head, and/or feather loss on ventral neck and body.
Biopsies
Skin biopsies were processed routinely for histopathology (Labtest, Auckland, NZ), with 3 sections of 6m stained with HE and examined under light microscopy (Leica 115
DM 1000, Leica Microsystems, Singapore). Images were taken using a camera mount for microscopes (Leica EC3, Leica microsystems, Singapore). Mites were only counted if sections of bodies [idiosoma] were visible (i.e., if only a leg was seen it was not counted). Mites were often visible across multiple sections, and easily identified as the same individual based on size, location, and surrounding structures. 120
These were not double counted.
Recovery and processing of mites
Frozen stored skin samples were defrosted and examined under a dissecting microscope. Mites were recovered by placing skin samples on a glass slide, adding a
84 drop of Hoyer’s medium, and dissecting out whole specimens using microtools 125
(Minitools, Entomology Supplies, Australia). Each slide was examined by light microscopy after clearing, with images and measurements for later taxonomic purposes taken of each mite.
Serial section examination
The distribution of mites, and mite abundance across different regions of the body, 130 was investigated in a single RCP with mange that was found dead in October 2012 on
Tiritiri Matangi Island. A set of serial sections was obtained using a 3mm punch biopsy from the frozen, archived carcass. Sections of skin (n=10) were taken from the cheek, ventral and dorsal neck, back, ventral and dorsal aspect of the patagium, lateral body wall over the ribs, abdomen, cloaca and the skin over the tibiotarsus. 135
These were processed as described above for other frozen skin sections.
Mite identification
Mites were identified to subfamily and species level based on the criteria described by Fain and Elsen (1967) for the Knemidokoptinae (previously Knemidocoptidae but subsumed as a subfamily in the Epidermoptidae by Mironov et al. 2005). Mites of 140 the Knemidokoptinae are small (<580um for females), with short legs and globose or oval bodies, and lacking spines (Fain and Elsen 1967). Unlike many mite species, adult females of the Knemidokoptinae are larviparous, so larval forms are frequently evident in gravid females. Only the female of P. janssensi has been described. Key characteristics that differentiate female P. janssensi from the similar K. pilae are 145 tarsal appendages on all legs and uninterrupted dorsal striations on the idiosoma.
85 The male, larva and nymph of P. janssensi are currently undescribed, although from our observations they share many features in common with the other species within the Knemidokoptinae. Larvae have six legs, nymphs have eight legs and lack genitalia, and the adult male has characteristic genitalia including an aedeagus. 150
Occasionally, mites enclosed in the cuticle of the previous stage were encountered, in which case the more advanced emergent form was recorded.
BFDV screening
Feathers and blood were screened by PCR for BFDV. The iGenomic blood DNA extraction kit (Intron Biotechnology, Gyeonggi-do, South Korea) was used to extract 155
DNA from blood and feather samples according to the manufacturer’s instructions.
PCR was performed using 4µl of the extracted DNA, KAPA blood PCR kit Mix B (KAPA
Biosystems, USA), and primers that target a ~605bp region of the replication associated gene as previously described (Julian et al. 2012, Julian et al. 2013).
Statistical methods 160
All statistical analyses were performed in “R studio” version 0.98.501, using Odds
Ratios (OR) for measures of association, and two-tailed fisher’s exact test for measures of significance where any one category was less than five, or chi-squared tests where all categories were greater than 5. The density histogram was also produced in “R studio” version 0.98.501. 165
86 5.3 Results
Sampling
In total, 142 birds were caught and sampled between 2012-2013 on Tiritiri Matangi 170
Island and Hauturu-o-Toi/LBI. Skin biopsies examined for mites were obtained from
135 birds.
Normal RCP skin
There were 21 skin samples without mites or other changes. Normal RCP skin had a thin epidermis, 1-2 cell layers thick, with a basket-weave keratin layer of 3-5 layers, 175 and minimal cellularity to the dermis (Figure 5.1a). Focal aggregates of predominantly lymphocytic cells were seen in four biopsies considered to be normal avian skin, akin to mural lymphoid nodules reported in birds (Carlson and Allen
1969).
Feather loss 180
During the period in which skin biopsies were obtained (2012-2013), feather loss was observed in 47/97 (48.5%, 95%CI: 38.2-58.8%) of RCP on Tiritiri Matangi Island, and
0/45 (0%, 95%CI: 0-7.9%) of RCP on Hauturu-o-Toi/LBI. Overall prevalence of feather loss for both islands was 47/142 (33.1%, 95%CI: 25.4 - 41.5%). Feather loss ranged in severity from mild periocular changes, through to loss of most feathers on the head, 185 ventral neck and sternum (Figure 5.2a and b). All birds with >50% feather loss on the head were similarly affected on the ventral neck and sternum, thereby satisfying the criteria for Grade 3 feather loss. Encrustations of the legs, beak or cere were not observed in the RCP examined. In some birds, featherless areas were mildly scaly,
87 pigmented and/or lichenified. Moderate to severe hyperkeratosis or encrustations 190 were not observed in any individuals. There were eight individuals with feather loss in which mites were not found, however there were histopathological changes
(hyperkeratosis and acanthosis) suggestive of mange.
Figure 5.2a and 5.2b: Red-crowned Parakeets with feather loss on Tiritiri Matangi 195
Island. Female with grade 3 feather loss affecting >50% of head, and mild scale formation (5.2a), and male with grade 3 feather loss, ventral neck and keel severely affected (5.2b).
Histopathological findings - general
Prevalence and range data for mites found in skin biopsies are presented in Table 200
5.2. Mites were significantly more likely to be located intrafollicularly versus extrafollicularly (p <0.0001, z-test). Mite numbers in skin biopsies were strongly
88 positively skewed in the density histogram (Figure 5.3). Mite numbers in biopsies correlated to severity of feather loss, with the odds of having Grade 3 feather loss significantly associated with >3 mites present in a skin biopsy (OR = 6.2, 95%CI = 205
1.74-22.05, p=0.004, Fisher’s exact test).
Table 5.2: Prevalence data for P. janssensi mites in skin biopsies, including locations of mites found, and range of total mite numbers observed in section. Mites overall indicates the percentage of skin biopsies (n=135) with mites observed, whereas intrafollicular and extrafollicular mites are expressed as a percentage of mite positive 210 biopsies (n=79).
x/n Prevalence 95%CI Range (biopsies)
Mites overall 79/135 58.5% 49.7-66.6% 1-11
Intrafollicular mites 68/79 86.0% 76.5-92.8% 1-9
Extrafollicular mites 37/79 46.8% 35.5-58.4% 1-6
Figure 5.3: Density histogram for mite counts from skin biopsies of Red-crowned
Parakeet on Tiritiri Matangi Island and Hauturu-o-Toi/LBI 2012-2013. 215
89 Hyperkeratosis and acanthosis were present in the majority of cases where mites or feather loss were noted, and were significantly associated with these explanatory variables (Table 5.3).
Table 5.3: Odds Ratios, 95%CI and Fisher’s exact p-value for relationships between presence of combined hyperkeratosis/acanthosis and mites in biopsy, feather loss, 220 or feather loss and mites combined.
Outcome of hyperkeratosis/ acanthosis OR 95%CI p-value
Mites in biopsy 8.93 4-19.91 <0.001
Feather loss 5.48 2.51-11.95 <0.001
Feather loss and mites in biopsy 31 9.23-104.14 <0.001
Inflammatory changes in the dermis were found in 72% (95%CI: 63.2-79.1%) of skin biopsies, ranging from mild perivascular to regionally extensive infiltrates involving predominantly mixed mononuclear cell lines with occasional granulocytic 225 dominance. There was, however, no significant relationship between these changes and the presence of mites (OR: 0.97, 95%CI: 0.46-2.07, p = 0.9).
Histopathological findings – mite stages and features
Different stages of intrafollicular changes were observed with mite infestation. All small mites of a size consistent with larvae, but two, were found at the epidermal 230 junction of the feather follicle in most cases, associated with varying responses of mild underlying acanthosis and hyperkeratosis, epidermal atrophy, and/or spongiosis. Individual mites appeared to be in the process of moving down the feather shaft towards the base of the follicle (Figure 5.1b). Mites of all sizes,
90 including gravid females, were observed in compact hyperkeratotic plugs or sheaths 235 at the base of feathers, overlying the dermal papilla and often widening the feather follicle itself, with acanthosis of the feather follicle epithelium (Figure 5.1c).
Mites of all sizes, including gravid females, were observed extrafollicularly, associated with compact orthokeratotic hyperkeratosis and acanthosis of varying severity. Some mites were found free-floating adjacent to skin sections. 240
Extrafollicular mites were also observed in small groups associated with a compact hyperkeratotic plug within the epidermis. A marked diffuse inflammatory infiltrate of both granulocytic and mononuclear cells was seen in the dermis on the rare occasion that mites were found penetrating the epidermis (see Figure 5.1d).
Sections of mites in skin were often longitudinal or transverse in nature, enabling 245 reasonable size and shape inferences to be made. Thus, in most cases, identification of life stage and mite family could be determined. Key characteristics typical of the
Knemidokoptinae were observed such as larval forms within adult females and cross sections demonstrating short legs relative to round bodies (Figure 5.1c). Rarely, the pretarsal stalk and pulvillus characteristic of P. janssensi females was observed, 250 which in combination with other features mentioned, enabled specific identification
(Figure 5.1e). Feces were regularly seen in sections behind mites, within mite bodies, or in hyperkeratotic crypts associated with mites. They appeared as round, pink clumps of keratin surrounded by a sheath (Figure 5.1b), and provided evidence of mites even in sections that did not contain mites. 255
91 Whole mite results
Results from examination of whole mites retrieved from frozen stored biopsies of seven birds are presented in Table 5.4, and demonstrate the dominance of P. janssensi. During dissection, mites were observed along the feather shaft, within 260 hyperkeratotic plugs in the epithelium, or at the base of feathers, consistent with the histopathology findings.
Table 5.4: Species, stage and total numbers of mites found in skin biopsies from the back of the head of Red-crowned Parakeets during cross sectional studies on Tiritiri
Matangi Island. The total number of mites found in the histopathology sections for 265 each individual are provided for comparative purposes under “n (histopath)”.
Procnemidocoptes janssensi Other mites
Gravid Non-gravid Larvae Total n (n) females females and (histopath) nymphs
Bird 1 6 11 12 29 5
Bird 2 3 4 6 13 2
Bird 3 1 4 10 15 8 Feather mite (2)
Bird 4 0 1 0 1 1
Bird 5 0 0 4 4 2
Bird 6 1 5 16 22 4 Epidermoptidae (1)
Bird 7 1 1 6 8 NA
Adult females had defining features of the Knemidokoptinae described in the methods (Figure 5.1f). Larval, nymph and non-gravid female mites were encountered in all sections, with features consistent with the Knemidokoptinae 270 generally, and the adult female P. janssensi more specifically. Features were
92 consistent across the mites found, supporting the identification by association despite lacking a formal description in the literature for the subadult stages of P. janssensi.
275
Figure 5.1: Stages of infestation with Procnemidocoptes janssensi in Red-crowned
Parakeet skin, including normal skin and a whole mite specimen. 5.1a) Normal skin
93 demonstrating the thin epidermis, basket-weave keratin layer, and minimal cellularity to dermis. HE. 5.1b) Widening of feather follicle due to presence of adult female mite at base (white arrow), with associated moderate acanthosis and 280 hyperkeratosis. Faeces evident (black arrow) proximal to the mite. Gnathosoma (A) visible with mouth-parts. HE. 5.1c) Adult female mite (black arrow) with chitinous (A) and developing (B) larval forms inside. Body shape and short legs of adult female support identification as knemidokoptinid mite. Larvae are also evident adjacent to the female mite (C). Mites are overlying the dermal papilla (white arrow) of the 285 feather follicle, with moderate surrounding acanthosis and hyperkeratosis. HE. 5.1d)
Marked mixed mononuclear and granulocytic diffuse dermal inflammation (white arrow) subjacent to mite penetrating the epidermis (black arrow). Marked acanthosis evident in adjacent epidermis. HE. 5.1e) Pretarsal stalk and pulvillus
(arrow) evident on second leg of adult female mite, definitive for P. janssensi 290 compared with other knemidokoptinids. Female mite has developing embryo (A).
HE. 5.1f) Cleared adult female mite with identifying features of P. janssensi, including chitinous larval form inside (A). Hoyer’s medium.
Serial sections of dead individual OBTM05
Examination of whole mites from serial sections of an opportunistically obtained 295 dead bird (OBTM05) with mange demonstrated the presence of large numbers of P. janssensi across the entire body (Table 5.5), despite clinical signs of feather loss being restricted to the head and ventral neck in this individual. Only a single male was found from 245 P. janssensi examined (0.4%, 95%CI: 0-2.3%). This, together with
94 other material obtained in this study provides the first specimens of the larva, 300 nymph and male of P. janssensi.
Table 5.5: Adult, nymph and larval forms of P. janssensi recovered from serial 3mm biopsy sections of skin from the body of a deceased Red-crowned Parakeet with mange (OBTM05).
Section Procnemidocoptes janssensi Other mites (n)
Gravid Non-gravid females Larvae Total females and nymphs
Cheek 2 8 4 14
Dorsal neck 7 11 10 28
Ventral neck 1 0 1 2
Abdomen 1 2 10 13
Back 3 4 6 13 Feather mite (1)
Lateral body 0 0 0 0
Dorsal patagium 6 2 26 34 Feather mite (1)
Ventral patagium 0 0 0 0
Cloaca 3 2 32 37
Tibiotarsus 3 7 2 12
305
Of all mites recovered and examined from skin sections, including frozen skin (n=92) and the serial sections (n=153), 245/250 (98%, 95%CI: 95.4-99.3%) were definitively identified as P. janssensi. The others were feather mites (Analgoidea; n=4) and a single unidentified epidermoptid, possibly Myialges (Promyialges) macdonaldi,
Evans, Fain and Bafort, 1963) which has been found on Hippoboscidae associated 310 with RCP (Heath 2010).
95 BFDV results
Two birds tested positive for BFDV between 2012-2013, however neither of these had feather loss. Mite counts from the biopsy taken from the back of the head were 315 zero and five for these two birds.
5.4 Discussion
The results of this study demonstrate a clear association between the presence of mites in skin biopsies, and histopathological changes of hyperkeratosis and acanthosis. These findings are consistent with a form of avian mange, although gross 320 manifestations of feather loss varied. This suggests a host-parasite-environment relationship that dictates the likelihood of developing clinical disease, as described with other mange syndromes (O'Brien 1999, Kolodziej-Sobocińska et al. 2014).
Relative abundance measures of mite burden revealed an aggregated, overdispersed pattern typical of macroparasite frequency distributions (Shaw et al. 1998), where 325 the majority of birds had 0-3 mites, and few birds harbored burdens of 4-11 per skin biopsy. Given almost all whole mites examined from a subset of birds with clinical mange were P. janssensi, the histopathological changes associated with knemidokoptinid mites at varying stages of infestation of both feather follicles and the skin, and the lack of gross, molecular, or histopathological evidence of a viral 330 aetiology for feather loss, the evidence strongly supports a causative role between this mite and the clinical mange observed.
96 Knemidokoptinae in wild and captive parrots 335
Procnemidocoptes janssensi has not been reported since the original description of a single non-gravid female mite from a Lovebird (Agapornis nigrigenis) that died several days after being imported to Belgium from Zambia in 1966 (Fain 1966).
Without knowing the original provenance of the bird (wild or captive) or how the bird was housed on arrival in Belgium, it is not possible to confirm whether the mite 340 was from Zambia, nor the likelihood the mite was from the wild. No confirmed reports of Knemidokoptinae in wild parrots were found during a literature search, with the majority of captive reports based on K. pilae in budgerigars (Lavoipierre and
Griffiths 1951, Oldham and Beresford-Jones 1954, Newton and O'Sullivan 1956,
Blackmore 1963, Rao et al. 1967). Knemidokoptinae in captive parrots other than 345 budgerigars are less commonly described, although are notable for the different distribution of lesions, compared with the classical beak, cere and leg lesions reported in budgerigars. For example, Knemidokoptinae morphologically different from K. pilae, although not described to species level, were found in association with feather loss of the head, neck, wings and thighs in 71% (95%CI: 52.5-84.9%) of a 350 group of captive RCP imported from Europe to Israel (Shoshana 1993). Mites described as K. pilae were also reported causing feather loss of the neck and ventrum in an Alexandrian parakeet (Psittacula eupatria) in England (Oldham and
Beresford-Jones 1954), head mange in a captive macaw (Ara sp) and unattributed signs in a parakeet (Psittacula krameri) in Holland (Jansen 1957). Further descriptions 355 of Knemidokoptinae (species unidentified) include those from a parakeet (species
97 unspecified) in Hawaii (Garrett and Haramoto 1967), and from captive Psephotus sp.,
Neophema sp., and RCP in Australia (Schultz 1978).
In New Zealand, K. pilae has been reported from captive budgerigars (O'Grady 1960), and a captive Yellow-crowned Parakeet (C. auriceps) with beak and cere lesions. This 360 latter record was initially ascribed to an RCP host (Bishop and Heath 1998b), but a later investigation and discussion with the aviary owner confirmed the error.
Knemidokoptinid mange (species unidentified) was reported in New Zealand in 19 wild-caught but captive-held Eastern Rosellas (Platycercus eximius) with scaly lesions of the head and legs (Gartrell et al. 2003). It is unknown if the birds acquired the 365 infestation in captivity or the wild. Morphological similarities between P. janssensi and K. pilae, coupled with a general lack of available expertise for morphological identification of parasites, could possibly have led to an under-reporting or misidentification of P.janssensi to date.
Clinical signs and pathology 370
Epidermoptidae and Knemidokoptinae appear to have adapted to specific microhabitats on hosts (Dabert et al. 2011), as well as to certain host species (Fain and Elsen 1967), and this typically governs the histopathological and clinical response that will be observed. Knemidokoptes pilae in Psittaciformes is generally found on unfeathered skin, often causing severe hyperkeratotic beak, cere and leg 375 lesions (Wade 2006), although a histopathological study of the stages of infestation noted presence of mites in feather follicles of the face (Yunker and Ishak 1957).
Knemidokoptes mutans causes similar encrustations restricted to the legs and feet in
Galliformes, as do K. jamaicensis and K. intermedius in Passeriformes (Dabert et al.
98 2011). Picicnemidocoptes laevis and N. gallinae are associated with de-pluming 380 syndromes or feather loss on the body of Columbiformes and Galliformes respectively (Fain and Elsen 1967, Rajabzadeh et al. 2008). The distribution of lesions on the head, neck and body in this study, coupled with the dominance of feather loss over formation of severe encrustations, suggests P. janssensi behaves differently from K.pilae, and, in fact, produces a clinical syndrome more akin to the de-pluming 385 mites N. gallinae and P. laevis.
The major histopathological findings from this study in association with mites were hyperkeratosis and acanthosis, which are non-specific changes that may be associated with a range of etiologic agents including parasites, viruses (Raidal 1995), and bacterial or fungal disease (Fletcher 2008). In this study, no other etiologic 390 agents were identified in skin sections on the basis of histopathological examination, and the observed changes were underlying or adjacent to parasites. Screening of blood and feathers from all birds for BFDV also ruled this out as a causative agent, supported by the lack of histopathological changes expected with this infection (Pass and Perry 1984, Fletcher 2008). A larger sample of BFDV positive birds would be 395 necessary to infer any relationship between infection with this virus and mite numbers.
The stages of mite infestation observed in biopsies gave insight into the mechanisms by which de-pluming or feather loss may occur, as well as the range of clinical presentations including carrier states. Mites found at the base of feather follicles are 400 likely to break the keratin bridges that play a significant role in attachment and strength of the feather (Fletcher 2008), as they move down the feather to this
99 position. Feathers were often absent when mites were located at the base of a feather, and the observed widening of the base may interfere with new feather growth. Consistent with other Knemidokoptinae, it appears P.jansenssi feeds on 405 keratin, as evidenced by the appearance of eosinophilic feces in section (Fain and
Elsen 1967).
Although a range of dermal inflammatory changes were observed in this study (e.g., mixed mononuclear and granulocytic infiltrates), they were not significantly correlated with the presence of mites. There were notable exceptions however, 410 where mites were clearly invading the epidermis and an influx of both granulocytic and mononuclear cell lines were observed. The absence of an association between dermal inflammatory changes and mites is consistent with other studies of
Knemidokoptinid mange in avian species (Mainka et al. 1994), and demonstrates the value of a large cross-sectional study that includes non-cases for comparative 415 purposes.
Methods and diagnostics
Obtaining biopsies from all birds regardless of clinical presentation was critical to detect the range of pathology related to the mite, the stages of mite infestation, and to provide baseline descriptions of normal skin. Given P. janssensi was 420 predominantly found in feather follicles, and does not form superficial encrustations containing large numbers of mites, skin scrapings used to detect Knemidokoptinae skin mites in other species and studies (Dabert et al. 2011) are unlikely to be as sensitive as skin biopsies, although this was not evaluated in this study. Detection issues for Knemidokoptinae using skin scrapes have been reported (Low et al. 2007), 425
100 highlighting the importance of obtaining the right sample based on the biology of the mite species involved. Shoshana (1993) successfully used 25 plucked feathers to look for mites in RCP with feather loss, based on the likelihood of finding mites in hyperkeratotic sheaths at the base of the calamus. This may have worked in the study reported here, however would not have allowed a description of the 430 histopathological changes associated with the presence of the mites, nor provided as many whole mites for identification as the biopsy method chosen.
There are recognized limitations to skin biopsies, including failure to obtain a representative skin sample for accurate diagnosis (Sleiman et al. 2013). A larger biopsy or multiple biopsies may have provided more information as to relative 435 abundance, ranges of pathology, and improved detection sensitivity for mites.
Biopsies give information on a single point in a temporally dynamic disease process; therefore within a population a range of stages and host responses are expected within the integument. It is possible to either miss the key lesion, or capture a developing or regressing lesion at a point where the inciting cause is no longer 440 present or evident (Sleiman et al. 2013). There were eight individuals with feather loss typical of mange in which mites were not detected, however absence of mites does not preclude the diagnosis of mange, as with mange in other species (Curtis
2012). Standardization of the biopsy site meant it was not always possible to sample the recommended leading edge of a lesion (Seltzer 2007), and thus the changes 445 observed in the biopsy may not reflect the full clinical spectrum in that individual.
However the serial section results from the single diseased bird demonstrated that mites can be distributed in feathered areas across the body of the bird, not only at
101 sites of clinical disease. Highest relative abundance included the nape of the head/neck, cloaca and dorsal patagium, further supporting the choice of biopsy 450 location. An additional limitation of biopsies is the need for appropriate chemical restraint, which, whilst necessary from a welfare perspective, adds additional logistical and personnel requirements to field studies. In this study however, the use of general anesthesia enabled a wide range of sampling and examinations to take place with minimal stress to the individual, a level of investigation uncommon in wild 455 bird studies of mange currently.
Whole mites obtained through stored frozen skin samples of affected birds were critical to accurately identify P. janssensi, assess relative abundance, and identify and enumerate other species or groups of mites present in RCP skin. This study provides a reminder of the diversity of parasites that may be found on wild individuals of any 460 species, and the need for histopathology of samples to determine if there are any deleterious host effects. Cleared specimens of P. janssensi, and unidentified epidermoptid and feather mites (possibly M. (P) macdonaldi, and Protalges sp. or
Pseudoalloptinus sp. recorded from RCP (Bishop and Heath 1998b, Heath 2010)) provided useful comparisons to differentiate mites in the histological sections 465 examined. Although features were observed in histopathological sections that indicated a mite from the Knemidokoptinae family (short legs, rounded bodies, larviparous adult females), it was extremely rare to find a feature such as a tarsal appendage on a female mite that would have enabled distinguishing them from K. pilae, the other common species in parrots. The fact that only one male mite was 470 found may reflect a species that is parthenogenetic, a technique bias, or a genuine
102 low relative abundance. Male mites of the Knemidokoptinae are considered to be difficult to detect (Oldham and Beresford-Jones 1954, Newton and O'Sullivan 1956).
Relationship between gross and microscopic pathology
Mite presence in histopathology sections did not always translate to feather loss or 475 skin disease, suggesting these individuals were either captured in the early stages of disease, during recovery, or were carriers of mites without clinical manifestation.
Carrier states, reservoir species, and a range of clinical disease is a common finding in populations exposed to and affected by mange causing mite infestations, both avian (Yunker 1955, Blackmore 1963) and mammalian (Kolodziej-Sobocińska et al. 480
2014). This reflects the complex host-parasite-environment relationship that governs the development of disease, with environment driving parasite abundance and longevity, as well as host health and immunity through resource availability. There is no information on the breeding biology and ecology of P. janssensi, and it can only be assumed that, like other knemidokoptinids, it spends its entire life cycle on the 485 host and requires close contact for transmission. Likewise, it is unknown whether
RCP are natural hosts for this parasite species or are infected through spill over events from sympatric species such as the Eastern Rosella or even other avian families. This host-parasite relationship may influence the severity of the host response and the likelihood of adaptive changes or host resistance to the parasite. 490
5.6 Conclusions
This is the first study to investigate a knemidokoptinid mange outbreak in wild parrots, providing histopathological evidence of the stages of infestation as well as
103 guidance on methods for identifying mites in sections. It reinforces the value of collecting representative biopsies from clinically affected individuals and controls of 495 the same species when investigating skin disease in wildlife populations. Wildlife disease research is increasingly common, however, particularly where parasites are involved such studies may be hampered by a lack of available expertise in morphological identification pending the more widespread availability of molecular tools. Given this study reports on relative abundances from a single biopsy from each 500 host, the high prevalence reported at a single site would suggest that most, if not all,
RCP at the study sites are likely to carry mites. Further studies should focus on the epidemiology of disease expression, and the biology of the mites themselves. We also recommend targeted sampling of all wild parrot species in New Zealand to determine the host species and geographic distribution of P. janssensi, including 505 remote islands that may provide insight into whether the mite is a native parasite of parrots in this region, or has potentially been introduced to the wild by either reintroduction programs using captive birds, or sympatric exotic parrots. It is currently unknown what impact this parasite may have on the conservation status and long-term viability of RCP and the broader parrot community in New Zealand. 510
Arguably, given the likely increased expression of disease in relation to adverse environmental and resource conditions affecting host immunity, as demonstrated in other mange studies, prevalence of this condition in RCP may act as a barometer for overall health and local environmental conditions in wild populations at a given time.
104 515
CHAPTER 6:
EPIDEMIOLOGY OF MANGE IN WILD RED-
CROWNED PARAKEETS
520
525
This chapter is modified from the following paper prepared for submission:
Jackson, B., Heath, A., Harvey, C., Holyoake, C., Jakob-Hoff, R., Varsani, A.,
Robertson, I., and K. Warren. 2014. Epizootic mange in wild Red-crowned Parakeets; seasonal trends and risk factor analysis. 530
105 6.1 Introduction
Mites (Acari) are a highly diverse and abundant group infesting a wide range of hosts, and a variety of locations on these hosts including fur, feathers, skin and mucous membranes (Krantz and Walter 2009). In many cases they are clinically insignificant, however certain species including those from the families Sarcoptidae 535 and Psoroptidae, and subfamily Knemidokoptinae, can cause mange with significant morbidity or mortality. This may lead to conservation impacts in wildlife (Carlson et al. 1982, Pence et al. 1983, Gaudioso et al. 2009) or economic losses and welfare issues in livestock and domestic species (O'Brien 1999, Losson 2012). Presence and severity of mange is governed by a variety of factors relating to the host (species and 540 health status), environment (conditions that favour parasite biology or affect host health), and the parasite itself (biology and strain virulence). Subclinical carriers or reservoirs often form a critical element in the epidemiology of mange (Kolodziej-
Sobocińska et al. 2014).
Mange epizootics, the outbreak of disease above a baseline enzootic level or 545 absence (Thrusfield 2007), have been reported in a variety of mammalian wildlife species including Western Grey Squirrels (Sciurus griseus griseus) (Carlson et al.
1982) and Iberian Ibex (Capra pyrenaica) (León et al. 1999). The major cause of mange in wild avian species results from infestation with mites of the family
Epidermoptidae (Gilardi et al. 2001), predominantly the subfamily Knemidokoptinae 550
(Dabert et al. 2011). However in wild birds reports of knemidokoptinid mange are more commonly case reports or series (Schulz et al. 1989, Jaensch et al. 2003, Miller et al. 2004, Holz et al. 2005, Dabert et al. 2011), or prevalence reports based on
106 clinical signs with only a subset of diagnostics in clinically affected birds (Mainka et al. 1994, Pence et al. 1999). Only one study has examined multi-year prevalences, 555 population scale impacts and risk factor analyses, with diagnostic testing of cases with no lesions to detect potential carriers (Low et al. 2007). The literature is dominated by reports in passerines, which may reflect a true over-representation of these species in mange epizootics, or the accessibility and visual prominence of these species in peri-domestic environments. 560
Knemidokoptinid mange has a range of host effects that appear to be driven by the mite species involved, with some mite species of showing host family specificity.
Thus K. pilae is only reported from psittacines, and predominantly the captive budgerigar (Melopsittacus undulatus), where it causes severe hyperkeratotic encrustations of featherless areas such as the beak, cere, and less often legs and feet 565
(Oldham and Beresford-Jones 1954). Neocnemidocoptes gallinae and
Picicnemidocoptes laevis cause a form of “depluming itch” leading to feather loss on the body in poultry and pigeons respectively (Fain and Elsen 1967), whilst K. jamaicensis, K. intermedius and K. mutans all cause hyperkeratotic encrustations of the legs and feet in a range of wild birds and poultry (Fain and Elsen 1967). 570
Knemidokoptes jamaicensis has been reported from 37 species across 13 passerine families (Dabert et al. 2013), representing a wide geographic and host species range compared with other knemidokoptinids such as K. pilae. However the growing body of research into these parasites coupled with the availability of molecular tools for barcoding may expand reported host species and geographic ranges, as well as 575 revealing cryptic host specificity (Whiteman et al. 2006).
107 Here we present findings from an epizootic of mange (characterised by feather loss) during a 2-year cross-sectional study of health and disease in a reintroduced, free- ranging population of Red-crowned Parakeets (RCP, Cyanoramphus novaezelandiae) on Tiritiri Matangi Island in the Hauraki Gulf of Auckland, New Zealand. A 580 comparative cross-sectional survey was conducted on a remnant wild population of
RCP on Hauturu-o-Toi/Little Barrier Island, also in the Hauraki Gulf. Histopathological results from skin biopsies during the second year on Tiritiri Matangi Island supported the diagnosis of ectoparasitic mange due to the knemidokoptinid mite
Procnemidocoptes janssensi (Chapter 5). This mite has only been described once, 585 from a lovebird (Agapornis nigrigenis) that died several days after being imported to
Belgium in 1966 (Fain and Elsen 1967). We provide the first epidemiological analysis of mange in a wild parrot species globally, including seasonal and locational trends and risk factor analysis.
6.2 Methods 590
Captures
Between 2011-2012, RCP were captured using standard mist nets (Avinet, USA) in
March and September of each year on Tiritiri Matangi Island (S36˚ 36' 2", E174˚ 53'
24"), to coincide with post-breeding and pre-breeding periods. RCP were reintroduced to Tiritiri Matangi Island during the 1970’s from a captive-bred 595 population (Dawe 1979). A comparative survey was conducted in June 2013 on
Hauturu-o-Toi/Little Barrier Island (LBI, S36˚ 12' 3", E175˚ 4' 54.84"), where a remnant wild population persists. Surveys took place over 1-2 week periods, and aimed to capture 65 birds to provide 95% confidence of detecting disease if it was
108 present at or above 5% prevalence in the population, assuming a test sensitivity of 600
0.9 and specificity of 1. A minimum target of 33 birds was set to detect disease above 10% prevalence with 95% confidence, using the same test characteristics.
RCP were placed in single-use cotton bags (Prospectors Earth Sciences, Australia) and taken to a field processing station where they were anaesthetised using isoflurane delivered via a zero dead space anaesthetic circuit (AAS, Sydney, Australia). Blood 605
(0.1-0.2ml) was collected from the medial metatarsal vein with a 1ml syringe and 25g needle, placed in a microhaematocrit tube for analysis of total protein (TP) and packed cell volume (PCV), and two fresh smears made using dedicated slide spreaders to provide white blood cell absolute and differential estimates. Following observed feather loss in 2011, a 3-5mm skin biopsy was taken from the back of the 610 head at the junction with the neck in all RCP in 2012, as well as on Hauturu-o-Toi/LBI, using fine sharp scissors and ensuring inclusion of at least one feather in the biopsy site. The site was chosen to represent an area commonly affected (the head), also in a location that was easily standardised across birds and would be less exposed or at risk of dehiscence during healing. Biopsy sites were closed with a single suture of 4-0 615 braided absorbable suture material (Polysyn, SVS, Auckland) and betadine ointment applied. A physical examination was performed including body condition scoring and body weight using digital scales (Wedderburn, New Zealand), metal and colour bands fitted (Department of Conservation, New Zealand), standard morphometric measurements obtained, and a series of photos of the head, ventrum and dorsum 620 taken, before the bird was recovered and released. No bird was re-sampled during a sampling session, however recaptures were sampled at subsequent sessions.
109 Laboratory testing
Skin biopsies were processed routinely for histopathology (LabPlus, Auckland, NZ), with three sections made of each sample, stained with HE and examined by light 625 microscopy for the presence of mites. Mites were counted according to the methods described in Chapter 5. In brief, a mite was only counted if a section of body
[idiosoma] was present. Identifying features of the mite and surrounding tissue prevented double counting across the 3 sections examined per bird.
PCV and TP were analysed in the field within 12 hours of collection using a 630 microhaematocrit spinner and a refractometer. Examination of blood smears for white cell indices were conducted by experienced avian haematologists at Gribbles
Veterinary Pathology (Auckland, New Zealand) and New Zealand Veterinary
Pathology (Auckland, NZ), using published methods (Samour 2000). Slides were excluded if smear quality prevented accuracy in white blood cell estimates. 635
Definitions and scoring
The case definition for “mange” in this chapter was the observation of feather loss, based on the findings from Chapter 5 and the correlation between feather loss of this nature with mites and histopathological changes supportive of a diagnosis of ectoparasitic mange. Therefore “mange” in this chapter specifically indicates birds 640 that have clinical changes associated with mite infestation. Comparatively, the case definition for “mites” in this study is any bird with a mite found in its skin biopsy, regardless of feathering.
110 Mange was recorded on a binary scale (Y/N), and then categorized on a scale of 0-3, where 0 = “no feather loss”, 1 = “mild patchy feather loss”, 2 = “patchy feather loss 645 with coalescing areas <50% of head”, and 3 = “patchy feather loss with coalescing areas >50% of head, that may include neck and keel”. There were no cases of feather loss on the neck/keel that did not include >50% feather loss on the head. No assessment of plumage condition was undertaken.
Biopsy samples were considered positive if one mite was found (i.e., a binary scale 650 for “mites”), and were then grouped in the following categories of according to the total number of mites observed across all sections examined: “0 mites”, “1-2 mites”,
“3-6 mites”, “>6 mites”. Mite numbers were also examined as a continuous variable to determine mean mite counts for comparisons.
Body condition was subjective and assessed by the same observer based on muscling 655 over the keel as reported for black cockatoos (Calyptorhynchus spp.) (Le Souef 2012) as well as fat stores over the lateral body wall. Categories were “emaciated”, “thin”,
“good”, and “very good”. These were also converted to a binary scale, with “normal” describing categories “good” and “very good”, and “abnormal” describing categories
“thin” and “emaciated”. 660
RCP are sexually dimorphic based on size/weight and beak characteristics (Sagar
1988). In this study, DNA sexing complemented by beak width and length measurements were used to assign sex.
665
111 Statistical analyses
Descriptive statistics including prevalences and 95% CI were calculated in excel and
“R studio” version 0.98.501 for mange, grade of mange, presence of mites, total number of mites, PCV, TP, estimated white cell counts (EWCC), and absolute lymphocyte, monocyte, eosinophil and heterophil counts. Analyses were repeated to 670 infer potential differences between sexes, sampling sessions and locations.
Dichotomous outcomes evaluated in this study were mange and mite presence in skin biopsies. The effect of different explanatory variables were analysed against these outcomes, including mange and mites (when the other was the dependent variable), total mites (categorical and continuous), severity of mange, location, 675 season, sex, weight, body condition, PCV, TP, EWCC, and absolute counts of lymphocytes, monocytes, eosinophils, and heterophils.
Measures of association were examined for nominal data (sex, location, season, year) and ordinal data (body condition, grade of mange, total number of mites in skin biopsy) using the Odds Ratio (OR) with 95%CI calculated in Epitools (Sergeant 680
2015). For ordinal data, OR were calculated against the group with the lowest attack rate (i.e., the lowest prevalence of the outcome of interest).
Measures of difference for nominal data were assessed by the chi-square test, or, where sample size in any one category was <5, the two-tailed Fisher’s exact test.
Continuous data (body weight, EWCC, individual white cell counts, mite counts) were 685 grouped by outcome and the difference of means determined by the two-tailed t-
112 test for normally distributed data, and the Kruskal-Wallis rank sum test for non- parametric data (to overcome issues of tied data).
All measures of association and difference were analysed for the entire dataset, then by location, sex, year, and individual sampling sessions (where sample sizes 690 permitted). Body weight was only analysed within the sexes, as RCP are sexually dimorphic with males weighing more than females.
Multivariate analysis for mange
Of the 229 birds examined for mange, 94 were removed as no biopsy was performed on these individuals, leaving 142 observations. A further 17 records were removed 695 because of missing haematology (n=15) and weight (n=2) data, leaving 118 observations for the final dataset for analysis. Variables for the analysis included sampling session, weight, sex, body condition (binary), mite count, and absolute counts of lymphocytes, monocytes, eosinophils, and heterophils. A logistic regression model was built manually from those variables that had Wald test p- 700 values less than 0.25, terms eliminated in a backwards stepwise manner, and interactions tested. Akaike information criterion (AIC) values were used to provide a relative evaluation of candidate models, in order to select the final model. These analyses were performed in “R studio” version 0.98.501.
6.3 Results 705
Relationship between mange and mites
For all birds and both islands combined, the odds of mange was significantly higher if one or more mite(s) were present in skin biopsies (OR=5.02, 95%CI: 2.2-12.7,
113 p<0.0001). This relationship remained for Tiritiri Matangi Island alone (OR=3.1,
95%CI: 1.3-8.4, p=0.017). No birds were observed with mange on Hauturu-o-Toi/LBI 710 despite a prevalence of mites in skin biopsies of 33% (95%CI: 19.6-49.5%), therefore a relationship between mites and mange was not found on this island at this time.
There was an association between the total number of mites per skin biopsy (a relative abundance measure) and mange across both islands (Table 6.1). The odds of mange was significantly higher in birds with greater than three mites in a biopsy. 715
Table 6.1: Logistic regression model showing the relationship between total mites in biopsy and the odds of mange.
Variable Mange No mange Positive Coefficient (SE) P OR (n) (n) (%) (95% CI)
Intercept -1.7918 (0.3819) <0.0001
1-2 mites 12 29 29.27% 0.9094 (0.5135) 0.08 2.48
(0.92-7.03)
3-6 mites 14 12 53.85% 1.8718 (0.5533) 0.0007 6.5
(2.25 – 20.07)
>6 mites 10 2 83.33% 3.4965 (0.8583) <0.0001 33
(7.3 – 241.9)
Seasonal and locational trends of mange (including severity), and presence of mites
Mange presence and severity 720
Prevalence data for mange across years, seasons and locations is presented in Table
6.2. The odds of mange were significantly greater on Tiritiri Matangi Island in 2012
114 compared to 2011 (OR=10.74, 95%CI: 4.5-25.62, p<0.0001). Although mange prevalence did not change significantly between March and September in 2012
(OR=1.52, 95%CI: 0.67-3.46, p=0.32), the severity of mange did. The odds of grade 2 725 or 3 mange was greater in September 2012 than March 2012 (OR = 3.04, 95%CI:
1.15-8.02, p=0.02). This significance remained when examining grade 3 mange against all other categories combined (0-2) between September 2012 and March
2012 (OR = 9.45, 95%CI: 1.17-75.99, p=0.028). No birds were observed with mange on Hauturu-o-Toi/LBI, therefore the odds of mange on Tiritiri Matangi Island across 730 both years were significantly higher than Hauturu-o-Toi/LBI (OR=38, 95%CI: 2.3-
628.01, p<0.0001).
Table 6.2: Prevalence of mange (Y/N) in Red-crowned Parakeets on Tiritiri Matangi
Island and Hauturu-o-Toi/Little Barrier Island across years and seasons.
Location Season/year Prevalence 95% CI n
Tiritiri Matangi Island April 2011 7.60% 2.1-18.2% 53
Sept 2011 8.80% 1.9-23.7% 34
March 2012 42.1% 26.3-59.2% 38
Sept 2012 52.5% 39.1-65.7% 59
2011 8.1% 3.3%-15.9% 87
2012 48.5% 38.2%-58.8% 97
2011-12 29.4% 22.9-36.5% 184
Hauturu-o-Toi/LBI June 2013 0% 0-8.4% 45 735 Mite presence and abundance
Mite presence data was only available for 2012 on Tiritiri Matangi Island, and 2013 on Hauturu-o-Toi/LBI, and is presented in Table 6.3. A locational and seasonal effect was noted, with the odds of having mites significantly higher on Tiritiri Matangi
115 Island than Hauturu-o-Toi/LBI (OR=4.64, 95%CI: 2.13-10.12, p<0.0001), as well as 740 being higher in September 2012 versus March 2012 on Tiritiri Matangi Island
(OR=2.79, 95%CI: 1.12-6.95, p=0.02).
Relative mite abundance (total mites per biopsy) was significantly higher in
September 2012 (x = 4.9, 95%CI: 4.1-5.8) versus March 2012 (x = 1.8, 95%CI: 1.3-2.3) on Tiritiri Matangi Island (p< 0.0001), as well as between September 2012 on Tiritiri 745
Matangi Island and June 2013 on Hauturu-o-Toi/LBI (x= 2.1, 95%CI: 1.3-28, n=42)
(p=0.0007). However the difference of means was not significant between March
2012 on Tiritiri Matangi Island and June 2013 on Hauturu-o-Toi/LBI (p=0.446).
Table 6.3: Prevalence of mites (Y/N) in Red-crowned Parakeets on Tiritiri Matangi
Island and Hauturu-o-Toi/LBI, across seasons and years. 750
Location Season/year Prevalence 95% CI n
Tiritiri Matangi Island April 2011 Not sampled NA NA
Sept 2011 Not sampled NA NA
March 2012 57% 39.5-72.9% 37
Sept 2012 79% 65.6-88.4% 56
2011 Not sampled NA NA
2012 70% 59.5-79.0% 93
Hauturu-o-Toi/LBI June 2013 33% 19.6-49.5% 42
116
Effect of sex on mite presence and mange (including severity of mange and total counts of mites in skin biopsies)
Mange (Y/N) and severity
No significant difference was detected for the outcome of mange when analysed by 755 sex, until analysis by season. The odds of having mange varied between the sexes on
Tiritiri Matangi Island in 2012, with males significantly more frequently affected than females in March (OR=6.26, 95%CI: 1.37-28.5, p=0.02), whereas females were more likely to have mange than males in September (OR=3.27, 95%CI: 1.13-9.51, p=0.03).
No significant relationships were found when analysing severity of mange (including 760 combining grade 2/3) by sex across years and individual sampling sessions.
Mites (Y/N) and abundance
Females were significantly more likely to have one or more mites in biopsies than males across both islands (OR=3.37, 95%CI: 1.63-6.96, p<0.0001) and for 2012 on
Tiritiri Matangi Island (OR=2.79, 95%CI: 1.1-7.09, p=0.03). However this significance 765 disappeared in March and was borderline in September for 2012 (OR = 4.76, 95%CI:
1.13-20.12, p=0.05).
A locational effect was detected for the relationship between sex and mean mite counts. Differences for mean counts were not significant between the sexes for
Tiritiri Matangi Island alone, however females had significantly higher counts on 770
Hauturu-o-Toi/LBI, and for both islands combined (Table 6.4).
117 Table 6.4: Results for mean mite counts in biopsies of Red-crowned Parakeets by sex and location
Males Females Kruskal Wallis p- Mites per biopsy range n range n value
Both islands, all years 1.6 0-9 71 2.7 0-11 64 0.003
Tiritiri Matangi Island Mar 2012 1.1 0-5 21 0.9 0-4 16 0.76
Tiritiri Matangi Island Sept 2012 3.2 0-9 26 4.4 0-11 30 0.16
Hauturu-o-Toi/LBI June 2013 0.2 0-2 24 1.3 0-5 18 0.004
Body weight and body condition
Relationship between body condition and mites or mange 775
The odds of mange were higher for birds as body condition decreased, with the most significant relationship between thin body condition and mange (OR = 16.57, 95%CI:
5.75–47.73), using very good body condition as the comparative factor. When grouping body conditions into a binary outcome, and then stratifying by location and season, mange remained highly significantly correlated with abnormal body 780 condition only for Tiritiri Matangi Island in September 2012 (OR = 27.3, 95%CI: 5.38-
138.42, p<0.0001).
Mites were also associated with thin body condition across the entire dataset
(OR=4.67, 95%CI: 1.64 – 13.30, p<0.0001), when compared to very good body condition. Following stratification by location and season, based on a binary 785 outcome for body condition, the relationship between abnormal body condition and mites remained only for Tiritiri Matangi Island in September 2012 (OR=27.33, 95%CI:
1.5-490, p<0.0001). No relationship was investigated between mites and body
118 condition on Hauturu-o-Toi/LBI, as all birds were in good to very good condition at that time. 790
Relationship between sex, body condition and mange or mites
There were differences in the relationship between mange and body condition when examined by sex. For females, the relationship between abnormal body condition and mange reflected that of the entire dataset, with the most significant association in September 2012 (OR=81, 95%CI: 6.45-1017.14, p<0.0001). For males, the 795 association between mange and abnormal body condition was significant when analysing the entire dataset (OR=6.6, 95%CI: 1.77-24.6, p=0.01), and Tiritiri Matangi
Island across both years (OR=4.69, 95%CI:1.25-17.61, p=0.03), however not in
September 2012 as noted for females (OR=6.38, 95%CI: 0.57-71.27, p=0.14).
There were no significant associations when examining abnormal body condition and 800 the presence of mites by sex.
Relationship between sex, body weight and mites or mange
Female body weights had a non-parametric distribution across the dataset versus males with normally distributed weights. Females with mange had significantly lower mean body weights on Tiritiri Matangi Island compared to females on Hauturu/LBI 805 for both years (p=0.02), for 2011 (p=0.02), for 2012 (p=0.04) and in September 2012
(p=0.005). No association was found in March 2012 for females (p=0.54), which correlates with body condition data. No significant associations were found between body weights of males with mange.
119 No relationships were established between body weights of males or females, and 810 the presence of mites.
Relationship between mites or mange and haematological parameters (PCV, TP, white cell counts)
Packed Cell Volume (PCV)
On Tiritiri Matangi Island, no significant differences were detected between mean 815
PCV values when analysed for outcome of mange or mites across seasons or years, and within seasons. Neither did analysis by sex reveal any significant differences.
PCV data was not available for Hauturu-o-Toi/LBI.
Total Protein (TP)
Females with mange had a significantly higher mean TP on Tiritiri Matangi Island 820
(p=0.004), as did all birds with mites (p=0.024). No relationship between mange and
TP was detected for males, and TP data was not available for Hauturu-o-Toi/LBI.
Estimated white cell count (EWCC)
No significant relationship was established between the mean EWCC and mange or mites, even when stratified by sex, location or season. 825
Lymphocytes
Females with mange had significantly lower absolute lymphocyte counts across the entire dataset (p=0.012), an association that only remained significant for 2012
(p=0.004) after examining by year and season. No relationship was detected between lymphocyte count and mange for males, or when examining the dataset for 830
120 both sexes. No relationships were detected between absolute lymphocyte count and presence of mites, regardless of location, sex, or season.
Monocytes
A significantly higher absolute monocyte count was found in both sexes with mange across both islands and all years (p=0.003), and all years on Tiritiri Matangi Island 835
(p=0.002). This difference remained highly significant for females on Tiritiri Matangi
Island when stratified by sex (p=0.0002), although disappeared for males (p=0.43).
No relationship was found between absolute monocyte count and mite presence.
Eosinophils
No significant differences were detected between absolute eosinophil counts when 840 analysed for the outcome of mange or mites across seasons or years, within seasons, and by sex.
Heterophils
Birds with mange had a significantly higher absolute heterophil count across both islands and seasons (p=0.019), which remained for females across the entire dataset 845
(p=0.02) although disappeared for males (p=0.33). Differences were not significant for absolute heterophil count when examined by year, season or sex beyond the entire dataset.
Conversely, birds with mites had significantly lower heterophil counts across both sexes on Tiritiri Matangi Island (p=0.013), and for males with mites on Tiritiri Matangi 850
Island (p=0.014).
121 Generalised linear model (GLM) for outcome of mange
Variables that were included in the full model based on the selection criteria were
“sampling session”, “mite count”, “body weight”, “sex”, “body condition” (binary) and “absolute heterophil count”. Results of model comparisons are shown in Table 855
6.5.
Table 6.5: Model contents of Generalised Linear Model (GLM) for mange in Red- crowned Parakeets on Tiritiri Matangi Island and Hauturu-o-Toi/Little Barrier Island.
The final model selected is highlighted in bold.
Model contents AIC Residual deviance
Full model (sampling session + mite count + weight + sex + body 93.082 75.082 condition + heterophil count) sampling session + mite count + body condition 89.682 79.682 sampling session + mite count * sex + body condition + mite 88.184 74.184 count + sex sampling session + sex + mite count + body condition 89.779 77.779
* indicates interacting terms 860
The final model selected included “sampling session”, “mite count”, “sex”, “body condition” and an interaction term between “mite count” and “sex”. Based on the final model, the odds of mange were 16.5 times higher (95%CI: 4.0 – 103.5) in birds with average/poor body condition, after adjusting for the effect of “mite count”,
“sex”, and “sampling session”. 865
122 Recaptured birds
Across the study, 17 birds were captured twice, and two birds were captured three times. All recaptures occurred at the same site of original capture. Two individuals were captured with mange (grade 1 and grade 2) in March 2012 that had returned to normal feathering by the time of recapture in September 2012, despite both birds 870 having mites in their biopsies. Both individuals were males. One other male was captured with grade 1 mange in March 2012, which had advanced to grade 2 mange by the time of recapture in September 2012.
6.4 Discussion
As described in Chapter 5, knemidokoptinid mites in RCP are significantly associated 875 with mange (i.e. feather loss and histopathological evidence of acanthosis and hyperkeratosis) of the head, neck and breast. Given the histopathological changes described in that chapter, they are considered the causal agent of clinical mange in this species. This study found that relative abundance (total mites per biopsy) influenced the likelihood of mange, indicating clinical signs are correlated with 880 increasing mite numbers, as has been described in common wombats infested with
Sarcoptes scabiei (Skerratt et al. 1999). A locational effect was noted for relative mite abundance, prevalence of mite positive biopsies, and mange, with the Tiritiri
Matangi Island population more affected than Hauturu-o-Toi/LBI. However this may have been confounded by the year of sampling and changes in resources and habitat 885 quality. On Hauturu-o-Toi/LBI no mange was noted, with a significantly lower prevalence of mites in biopsies than on Tiritiri Matangi Island. This may be indicative of the enzootic or baseline state of infestation, although given the biopsy only
123 reflects a very small section of skin, it is possible all birds harbour mites, with specific host or environmental conditions required for an outbreak of mange to occur. 890
An epizootic of mange was detected in 2012 based on annual prevalence data, with mite presence, total mite numbers per biopsy, and severity of mange all significantly increasing by the second sampling session in September 2012. Reasons for this outbreak may relate to the host, environment, or parasite, including synergistic interactions between multiple factors. The outbreak event was preceded by a La 895
Niña summer characterized by cool, wet conditions (NIWA 2012). Vegetation and climate variables were not investigated in this study, however La Niña conditions may have impacted flowering plants and resource availability, which in turn may have impacted host nutritional status and immunity. The biology of the mite in question is undescribed, however cool humid conditions favour off host survival in 900 other species of skin mites (Scott and Miller 2011), therefore the weather preceding the outbreak may have altered transmission factors. A longer term dataset would be necessary to infer climate effects on interannual trends in prevalence data.
Climate and resources may affect both host health and transmission factors. An outbreak of notoedric mange in western grey squirrels was attributed to multiple 905 factors including a failed mast crop and consequent nutritional stress, with poor resource availability driving dispersal and/or contacts between hosts thereby facilitating transmission (Carlson et al. 1982). These same factors may play a role in knemidokoptinid mange in RCP, as the genus is known to respond to masting events with increased breeding activity (Elliot et al. 1996), and RCP specifically have 910 significant inter-annual variation in reproductive capacity (Ortiz-Catedral and
124 Brunton 2010a). Masting events and resource availability may influence the expression of mange in RCP populations in a number of ways. Increased nutrition will influence host health and immunity, which may suppress the expression of clinical mange in the population. Also, younger birds are less commonly affected by 915 knemidokoptinid mange than older birds (Shoshana 1993, Mainka et al. 1994,
Jaensch et al. 2003), therefore the influx of clinically normal juveniles into a population during a significant breeding event may lower the prevalence of mange detected.
We hypothesise clinical mange will be less prevalent in years of increased resource 920 availability due to optimised host nutrition and immunity, combined with a higher proportion of juveniles in the population that are unlikely to demonstrate clinical signs. In the outbreak year, a combination of factors including a relatively aged population plus conditions leading to poor resource availability and favouring parasite persistence, may have led to greater susceptibility to, and expression of, 925 clinical mange. However as it is not possible to reliably age RCP beyond their first 6-
12 months, and this study did not investigate age, nutrition or immunity as risk factors, this hypothesis remains speculative and based on extrapolation from theoretical knowledge for mange in other species. Banding a significant proportion of the population and repeat sampling over a longer period would be the simplest 930 way of gathering age-related data (along with other benefits of long-term studies such as increasing statistical power through sample sizes).
Results from the RCP population on Hauturu-o-Toi/LBI are interesting as the population is remnant wild (compared to the introduced from captivity population
125 on Tiritiri Matangi Island), there are no known records of introduced exotic parrots 935 visiting the island, and the habitat composition may lead to a more dispersed population of RCP than on Tiritiri Matangi Island. These factors provide insight into the potential origins of the parasite found in both populations. It is unlikely a host- switching event may have occurred between exotic parrots and RCP on Hauturu-o-
Toi/LBI, given the rarity of visiting exotic parrots, coupled with the presumed need 940 for close contact based on the biology of knemidokoptinids generally. It is possible that infested RCP may have immigrated from Tiritiri Matangi Island, bringing with them a captive-derived parasite. However the findings suggest the parasite may be native to New Zealand generally and RCP specifically, with molecular biogeographic studies and surveillance of remote island RCP and sympatric parrots necessary to 945 resolve this question.
Birds sampled appeared to be in a period of relatively good health on Hauturu-o-
Toi/LBI based on the lack of observed feather loss and good to very good body condition, compared to the Tiritiri Matangi Island population. However historically the Hauturu-o-Toi/LBI population has had a high rate of BFDV infection detected 950 with clinical signs of feather loss in juvenile and adult birds (Ortiz-Catedral et al.
2009b). Interestingly, a post mortem case series from the year of BFDV detection in the Hauturu-o-Toi/LBI population, described intrafollicular mites of unknown species in 7/7 birds examined, associated with mange predominantly on the ventrum, as well as folliculitis and dermatitis (Brennan and Alley 2008). It is possible these mites 955 were the same as those found in this study, and that co-infection with BFDV and associated immunosuppression may augment the development of clinical mange.
126 The results from Hauturu-o-Toi/LBI also highlight the ongoing presence of mites in a normally feathered (at the time of sampling) population of birds, supporting the likelihood of an enzootic/carrier state between epizootics of mange. Sample sizes 960 were sufficiently large to be statistically confident of detecting feather loss if it had been present at a prevalence >10% (i.e. an outbreak level) in this population.
Risk factor analysis
Seasonal variability in the expression of mange between the sexes was detected, with females more affected in September 2012, and males more affected in March 965
2012. This may relate to breeding activity, which varies in intensity for both sexes at different times of the year, or may be a product of small sample sizes when analysing for sex at the level of sampling session. A seasonal effect on mange prevalence due to an unidentified knemidokoptinid mite was described in hihi (Notiomystis cincta) by Low et al .(2007) on Tiritiri Matangi Island, with the male sex bias likely related to 970 testosterone-driven immunosuppression during the breeding season. Females in our study were more likely to have mites without mange on Hauturu-o-Toi/LBI, where a presumed baseline enzootic state was occurring at the time of sampling. This may reflect a higher tolerance of females for mite burdens prior to clinical signs, although some caution must be taken with interpretation of a relative abundance measure 975 from a small skin biopsy. Given females spend the majority of time in contact with chicks in the nest, particularly young chicks; this could also be a parasite strategy to increase the likelihood of transmission. During the outbreak year the effect of sex on mite abundance disappeared, suggesting both sexes are similarly infested during these periods. 980
127 Mite presence in skin biopsies, as well as mange, was associated with declining body condition, particularly in September 2012 when severity of mange reached its peak.
In the final multivariate model, body condition was most significantly correlated with clinical signs of mange after adjusting for all other effects, indicating the strength of the association. An association between mange and body condition was not 985 detected in hihi with severe (presumed knemidokoptinid) mange (Low et al. 2007), however supplementary feeding in this species at this location may confound this data. Poor body condition is known to correlate with mange presence and severity in mammalian species (Skerratt et al. 1999, Oleaga et al. 2012), and ectoparasitism in general (Lehmann 1993). However whether body condition precedes mite increases 990 and signs of mange, or is a result of the parasitism, is unknown. More severely affected RCP in this study were observed to be pruritic, spending time preening, shaking and scratching whilst perched, and appearing highly irritated. This may have affected time budgets, including feeding, that may in turn create a negative feedback loop for body condition and increasing ectoparasitism. Future studies that 995 incorporate behavioural ethograms for mange affected and un-affected birds would help clarify this aspect.
Haematological changes were not significantly associated with mites or mange, except within the female cohort, where mange was significantly associated with an increased TP and monocytosis during the outbreak year. An increased TP may 1000 indicate a host inflammatory response due to increased globulins and monocytosis is observed during chronic inflammation in avian species (Harrison and Lightfoot 2006), both biologically plausible responses to parasitism in this study. The reason why
128 females were more likely to respond in this way is unclear. Reported reference range data for haematology of RCP did not detect sex differences in the haemogram of 1005 normal birds, therefore it is unlikely sex was a confounder for the differences detected in this study. The failure to detect haematological differences in birds with mites, as compared to those with mange, suggests changes in TP and circulating monocytes are associated with clinical disease rather than the presence of mites alone. 1010
Recovery from mange
Recapture of two birds that had recovered from mange, although still had mite positive biopsies, demonstrates that survival and recovery from knemidokoptinid mange is possible in RCP, although does not mean birds are no longer carrying mites.
In other species, knemidokoptinid mange lesions have been reported as chronic and 1015 progressive, and rarely go into remission without treatment (Pence et al. 1999), however Low et al. (2007) noted survival was not significantly affected in a closely monitored group of hihi with mange on Tiritiri Matangi Island. We cannot infer survival from this dataset, therefore it is unknown if significant mortality or morbidity occurred in other mange-affected individuals not recaptured during the 1020 study period. To determine the population impacts of this parasite, a study incorporating population size against prevalence data would be needed. A long-term dataset of this nature would also provide valuable information as to the capacity for population recovery post epizootic, as has been captured following high mortalities in coyotes affected by sarcoptic mange (Pence and Windberg 1994). 1025
129 Transmission factors
The biology of P. janssensi is unknown, although most members of the
Knemidokoptinae spend their entire lifecycle on the host, and require close contact for spread. It is therefore likely that the transmission dynamics are driven by host 1030 contacts, however a better understanding of the social dynamics of RCP will help determine if transmission is density-dependent, frequency-dependent, or both
(Ryder et al. 2007). This data is important to understand host-parasite equilibria, and whether population thresholds exist for parasite persistence (Lloyd-Smith et al.
2005, Ryder et al. 2007). RCP are not a gregarious parrot, forming pairs or small 1035 groups, with data from recaptures in this study suggesting a high degree of territoriality even on small islands such as Tiritiri Matangi Island. Key transmission points would include the nesting period and allo-grooming between family groups and pairs. Similar to hihi with mange (Low et al. 2007), RCP were observed to be pruritic and rubbing heads on branches. Depending on off-host survival of the 1040 parasite, branches may act as fomites, particularly in areas where birds congregate such as artificial watering points around Tiritiri Matangi Island.
It is also possible the RCP population on Tiritiri Matangi Island is at carrying capacity, with population density of this and other species more rapidly depleting resources in difficult years, as well as having higher host contact rates compared to Hauturu-o- 1045
Toi/LBI. Further research into the biology and ecology of the parasite, combined with host dynamics including social network models (Godfrey et al. 2010), would provide significant insight into the mechanisms for transmission and persistence of P. janssensi in RCP. The importance of clearly defining and linking parasite life cycles
130 with host behaviour is critical to a thorough understanding of the epidemiology of 1050 infectious diseases (Grear et al. 2013).
Several authors have attempted experimental infection with Knemidocoptes pilae
(Kutzer 1964, Kirmse 1966) with poor success. In one study, close contact between a heavily infected hen and cock bird did not result in clinical disease in the male bird, neither did housing a heavily infected hen with chicks produce clinical signs in 1055 younger birds over a 6 week period (Blackmore 1963). These studies led to the assumption that there are poorly understood factors, likely including nutritional status and condition of the bird, that will result in clinical mange when birds are exposed to knemidokoptinids (Fain and Elsen 1967). Outbreaks of mange in other species have been attributed to parasite mutations leading to increased virulence, 1060 facilitated (but not initiated) by host density (Pence et al. 1983). Without knowing the status of sympatric native and introduced exotic parrots such as the Eastern
Rosella, it is not possible to say what role these other species play as reservoirs for P. janssensi, although the presumed close contact required for transmission makes this less likely except in aviary conditions. 1065
Methods
The inclusion of a biopsy was critical to detect the mite species causing pathology as well as identifying subclinical carriers (see Chapter 5). Control birds with no evidence of feather loss provide important information to determine risk factors and host characteristics that drive the clinical expression of mange, a feature lacking in most 1070 population scale studies on knemidokoptinids (Pence et al. 1999). Single use bags for holding birds, and biosecurity measures (washing hands, changing gloves) between
131 handling birds were measures used to prevent parasite transmission between individuals. Carlson et al (1982) observed re-use of traps during a study on mange in squirrels may have contributed to increased transmission and the outbreak 1075 described. The importance of appropriate biosecurity during wildlife research to prevent disease transmission cannot be over-emphasised.
Management and control
Reported individual bird treatments for knemidokoptinid mange include ivermectin
(Pence et al. 1999, Koski 2002, Rosskopf Jr. 2003, Miller et al. 2004, Low et al. 2007), 1080 moxidectin (Shoshana 1993, Holz et al. 2005), and 0.1% trichlorofon solution
(Rajabzadeh et al. 2008). In severe cases repeated treatments may be needed until clinical signs resolve (Rosskopf Jr. 2003). Treatment for mange may be indicated for individual birds involved in captive breeding programs, or translocation and reintroduction events. This is especially important if the skin mite status of a 1085 recipient population is unknown, to prevent range expansion of the parasite pending further surveillance. However management of mange at a population scale using these medications is neither practical, nor necessarily appropriate unless population scale impacts have been determined. It is more important to understand the drivers of mange in RCP populations, to determine if management actions to mitigate 1090 parasite transmission and/or disease expression are possible at this level, for example if supplementary feeding or water points encourage congregation and host contacts.
132 6.5 Conclusions 1095
This study investigates an epizootic of mange due to infestation with P. janssensi in
RCP on Tiritiri Matangi Island, with a presumed enzootic state of infestation without clinical signs detected on nearby Hauturu-o-Toi/LBI. Body condition, sex, year and season were significant risk factors for expression of clinical signs of feather loss, suggesting there are ecological and host factors that contribute to the emergence of 1100 disease above baseline or enzootic levels, as well as inter-annual variation. Individual mortality rates and population scale impacts of this parasite are unknown, however recovery from mange was established in two recaptured individuals in this study. We strongly recommend further research into the epidemiology of disease and population impacts of ectoparasitic mange in RCP and sympatric New Zealand 1105 parrots; (i) to delimit mange and infestation with mites spatially and by host species thereby determining if biosecurity measures should be implemented to prevent spread to small/isolated populations such as the critically endangered Malherbe’s
Parakeet (Orange-fronted Parakeet, Cyanoramphus malherbi); (ii) to determine the need and scope for intervention or management at the population or individual 1110 scale, including for captive breeding and reintroduction programs; and (iii) to further understand the ecological risk factors that may relate to broader species conservation issues such as climate variability and habitat quality.
133 Chapter 7: 1115
Nesting study to investigate correlations
between reproductive success and nest
mites, knemidokoptinid mange, BFDV
infection and reproductive parameters
1120
This chapter is a modified version of the following paper prepared for submission 1125 to EMU (Austral Ornithology):
Jackson, B., Heath, A., Holyoake, C., Jakob-Hoff, R., Varsani, A., Robertson, I., and K.
Warren. 2014. Impact of haematophagous nest mites (Dermanyssus sp.) and knemidokoptinid mange mites on reproductive parameters and fledgling haematology in wild Red-crowned Parakeets (Cyanoramphus novaezelandiae). 1130
134 7.1 Introduction
Parrots, one of the most diverse and characteristic avian groups, are also one of the most imperilled, with 34.6% of species classed as threatened (Forshaw 2010). New
Zealand is home to eight endemic species of parrot, including the critically endangered kākāpō (Strigops habroptilus) and Malherbe’s Parakeet (Cyanoramphus 1135 malherbi), the endangered kākā (Nestor meridionalis), and the vulnerable kea
(Nestor notabilis) (BirdLife International 2013). Most species are actively managed through captive- or assisted-breeding programs, reintroduction and translocation of individuals, or management support such as provision of nesting boxes. The
Cyanoramphus genus, including the Red-crowned Parakeet (RCP, Cyanoramphus 1140 novaezelandiae), has been the subject of numerous translocation and reintroduction attempts in New Zealand with varying success (Dawe 1979, Ortiz-Catedral and
Brunton 2010b, Ortiz-Catedral et al. 2012). Tiritiri Matangi Island in the Hauraki Gulf of New Zealand is the site of an early translocation program of RCP in the 1970’s
(Dawe 1979). Removal of introduced mammalian predators and revegetation of the 1145 island subsequent to the reintroduction has likely contributed to the stable and growing population of RCP (Graham et al. 2013).
Red-crowned Parakeets are secondary cavity nesters like most parrots, but are known to explore a more diverse range of nest sites including seabird cliff burrows, the base of flax plants (Phormium tenax) and other ground nesting sites (Greene 1150
2003, Ortiz-Catedral and Brunton 2010a). This flexibility in nest site selection makes them a good species to introduce to restored habitat where old-growth forest and appropriate cavities may be limiting. Masting events that drive episodic and
135 significant increases in available food sources have been noted to stimulate increased breeding activity in members of the Cyanoramphus genus, including the 1155 production of multiple clutches and year-round breeding (Elliot et al. 1996). These nesting characteristics mean that RCP have the reproductive capacity to more rapidly restore numbers in response to stochastic events or sudden declines, compared with species with low reproductive potential such as the kākāpō.
Breeding success is critical to the establishment and persistence of a species in a new 1160 environment. The potential causes of breeding failure are diverse and include a variety of social factors such as conspecific and intraspecific aggression (Krebs 1998), ecological constraints such as availability of nest sites and environmental fluxes in resources (Burke and Erica 1998), genetic influences on fertility and hatching success
(Heber and Briskie 2010), diseases including parasites (Ewen et al. 2009), and 1165 predators. Thus, long-term and comprehensive studies are necessary to identify the range of possible causes of suboptimal breeding in a given population. An understanding of all these factors is critical, especially where species' management involves translocations/reintroductions or captive-breeding programs, as they can help to identify strategies that improve success and reduce population and program 1170 costs.
The impacts of disease on nestling survival have not been specifically studied despite the detection of Beak and feather disease virus (BFDV) in wild RCP (Ortiz-Catedral et al. 2010), with its highest mortality in juvenile parrots (Raidal 1995, Raidal and Cross
1995). Other disease processes that may affect reproduction and fledging success in 1175
RCP have not been studied, including the mite (Procnemidocoptes janssensi) that
136 was found to cause epizootic mange in the adult RCP population on Tiritiri Matangi
Island (Chapter 5 and 6). It is unknown what impact these mites may have on reproductive parameters in this species, or whether chicks of affected parents will develop clinical mange by the time of fledging. 1180
Haematophagous nest mites are known to affect reproductive parameters in a range of species globally (Møller 1990, Berggren 2005), and have been reported in wild RCP nests (Greene 2003) as well as in artificial nests provided for other species but used by RCP on Tiritiri Matangi Island (Stamp et al. 2002). Nest mites described from
Tiritiri Matangi Island include Ornithonyssus bursa (Berggren 2005, Ewen et al. 2009), 1185 and O. bursa combined with an un-described Dermanyssus sp. (Stamp et al. 2002).
Their impact on reproductive success in RCP remains unexplored. Nest mites such as
D. gallinae, O. bursa and O. sylviarum are considered major parasites in the poultry industry, and have a global distribution in wild birds and poultry (Owen et al. 2009,
Marangi et al. 2012). These mites have also been found on many species of domestic 1190 and wild birds in New Zealand (Bishop and Heath 1998b, Heath 2010).
Nest mites are directly pathogenic via haematophagy (blood-sucking). Dermanyssus gallinae has been shown experimentally to transmit Western Equine Encephalitis and
Fowl Pox viruses, as well as bacteria such as Coxiella burnetii (agent of Q fever) and
Pasteurella multocida (Valiente Moro et al. 2009). The mite is also implicated as a 1195 vector for major zoonotic pathogens including Salmonella sp. (Valiente Moro et al.
2007) and Chlamydia psittaci (Circella et al. 2011). Clinical signs of nest mite infestation in poultry and pet birds include irritation, anaemia, weight loss and decreased egg production (Circella et al. 2011). Nest mites form dense populations
137 at feeding trays and nest boxes (Cencek 2003), and are a public health issue in the 1200 poultry industry where they may cause dermatitis whilst feeding on humans (Orton et al. 2000).
This study is the first report on the relationship between disease risk factors such as the presence of nest mites, BFDV infection, and ectoparasitic mange of parent RCP, with outcomes including selected reproductive parameters, haematology and 1205 physical examination findings in RCP chicks. As previously identified (Ortiz-Catedral et al. 2013), the study species provides an accessible model from which to infer conservation threats for all members of the Cyanoramphus genus, and for New
Zealand parrots generally. Given the problems that cosmopolitan nest mites pose, the undescribed dermanyssid investigated here provides insight for conservation 1210 managers of other avian taxa in New Zealand and globally, including an evaluation of control and management options.
7.2 Methods
Species and Location
Red-crowned Parakeets are a small to medium sized sexually-dimorphic parakeet, 1215 where males are larger than females (Sagar 1988). They are generally found in pairs or small groups, with flocking usually observed only for key food sources (Sagar
1988). Tiritiri Matangi Island (S36˚ 36' 2", E174˚ 53' 24") is located in the Hauraki Gulf approximately 30km from Auckland, and is co-managed by the Department of
Conservation (DOC) and a community group, the Supporters of Tiritiri Matangi 1220
(Galbraith and Cooper 2013). It is an open sanctuary dominated by regenerating
138 young forest planted during a 10-year period from 1984, with patches of open grassy areas and limited old-growth forest (Rimmer 2009).
Nest box design and placement
In October 2011, 31 wooden nest boxes were placed on the island (Figure 7.1) in 1225 areas reflecting sampling of birds as part of a broader study by the author on health and disease in RCP. Boxes were made of macrocarpa (Cupressus macrocarpa) timber, measured 25cm x 20cm x 15cm, with a 6cm entrance hole, a wooden perch, an inlaid ladder on the interior leading to the entrance hole, and a hinged lid to allow observer access (Figure 7.2). Nests were filled with wood shavings (untreated pine, 1230
Pinus radiata), and were placed on trees or metal stakes (in areas of flax plants), approximately 1 metre above the ground.
Figure 7.2: Wooden nest boxes made of macrocarpa for Red-crowned Parakeets on
Tiritiri Matangi Island. 1235
139 In September 2012, 37 plastic nest boxes were placed on the island (Figure 7.1) in areas where birds were most likely to use them; an assumption based on a relative lack of available natural sites. Boxes were made with 30cm diameter light coloured
PVC (polyvinyl chloride) downpipe (to reflect light and heat). They were 50cm high, with a 6cm entrance hole and wooden replaceable perch (Figure 7.3a and 7.3b, and 1240
Figure 7.4). There was an inlaid hinged door on the side for access, and two wood ladders attached internally leading to the entrance hole. Nests were filled with wood shavings as for wooden nest boxes (Figure 7.5), and were placed on trees as for the wooden boxes. Plastic boxes were introduced as a preferred nest type for durability, and the ability to effectively disinfect for BFDV between seasons. 1245
7.3a 7.3b
Figure 7.3: Components of plastic nest box design for Red-crowned Parakeets on
Tiritiri Matangi Island. 7.3a – Hinged door with inlaid rim for waterproofing, 7.3b – removable wooden ladders leading to exit, with ventilation holes visible at the top of the nest box. 1250
140
Figure 7.4: Plastic nest box for Red-crowned Parakeets in situ on Tiritiri Matangi
Island.
Figure 7.5: Inside plastic nest box with wooden ladders visible and 8 eggs on wooden 1255 shavings used as nesting substrate.
141 Natural nests
Natural nests were identified by looking for nesting behaviour (mate feeding or entering/exiting of cavities), listening for chick or adult calls as they approached the 1260 nest, or looking for characteristic faecal piles at the entrance to nests. Natural nests were identified at different stages of the breeding cycle and therefore sample sizes for analyses of reproductive parameters vary.
Figure 7.1 – Map of Tiritiri Matangi Island showing location of nest boxes and 1265 successful nests, including plastic boxes (), wooden boxes (), and natural nests
(). Successful nests (chicks fledged) are filled in with black, with number of chicks fledged indicated by the number to the top right of the nest.
1270
142 Nest monitoring
All nest boxes were monitored once weekly from October 2012 until March 2013.
Natural nests were identified and monitored during the same period and interval.
For all nest types, observations were undertaken only when the male called the female off the nest to feed, and were completed quickly to ensure females could re- 1275 enter the nest undisturbed. Nest boxes were checked via hinged lids (wooden) or access doors (plastic). Natural nests were accessed by a variety of methods depending on location and type, including the use of a 4m borescope (EasyLife
Products, Hong Kong), where depth prevented safe access without disturbing the nest structure or brood. Clutch size, live young, unhatched eggs, and chick 1280 development were recorded at each check.
Parents were observed for signs of feather loss, known to be associated with ectoparasitic mange on the island (Chapter 5). As capture of parent birds for skin biopsy was not undertaken in this study, the case definition for mange was based on previous work where patchy feather loss on the head, neck and keel strongly 1285 correlated with presence of knemidokoptinid mites and mange pathology (Chapter
5). Therefore a positive mange case was defined by moderate to severe patchy feather loss on the head and/or neck region with exposed skin, as milder signs could be confused with the normal moulting process or other causes of feather damage.
Categories of mange-positive and mange-negative were used in later analyses. 1290
143 Nest mite collection and processing
Nests and chicks were examined for mites weekly, and representative samples from each nest site were collected from nest material or door hinges using a gloved hand 1295 and placed into 70% ethanol. These were cleared and mounted on glass slides in
Hoyer’s medium for identification. Dermanyssid mites were identified by reference to a standard text (Krantz and Walter 2009), as well as references listed in Roy and
Chauve (2007).
Chick examination and sample collection 1300
Approximately one week from fledging, or 30-35 days of age, chicks were sampled for various diagnostic tests, examined for general health and feather condition, and banded with a single metal band (DOC, Wellington, NZ). Chicks were removed from the nest as a group and placed in a holding container with fleece lining, processed individually and replaced into the nest afterwards. Blood (< 0.1ml) was collected 1305 from the medial metatarsal vein using a 25-gauge needle and 1ml syringe. Two fresh blood smears were made using a dedicated slide-spreader to maximise slide quality and accuracy of interpretation (Capitelli and Crosta 2013), and the remaining blood placed on Whatmann No.3 filter paper and air-dried. Growing feathers (2-5) were collected from the breast using forceps, and placed in 70% ethanol. Standard 1310 morphometric measurements were taken (wing length, tarsometatarsus length and beak width/length), chicks were weighed with digital scales (Wedderburn, Auckland,
New Zealand), and a standard set of photographs obtained.
144 Haematology 1315
Blood smears were fixed in methanol, then placed for 1 minute in a bath of
Leishman’s stain, pH 6.8 phosphate buffer, and May Grunwald stain diluted 50:50 with pH 6.8 phosphate buffer (Dacie and Lewis 1995). An area with uniform red blood cells was selected and an estimated white cell count (EWCC) performed by counting all white blood cells (WBC) in 10 fields at 40x objective, dividing by 10 to 1320 get an average, then multiplying by 2 to obtain the TWCC x 109/l (Samour 2000).
Differential counts were performed using oil immersion at 50x magnification, with each WBC encountered counted until reaching 100, to obtain a relative differential
(%). The absolute differential was calculated by multiplying the relative differential
(%) x the TWCC, to obtain a count x 109/l. 1325
Molecular testing for BFDV
DNA was extracted from feathers stored in ethanol and blood stored on filter paper, using the iGenomic blood DNA extraction kit according to the manufacturer’s instructions (Intron Biotechnology, South Korea). BFDV was screened for by PCR using 4µl of the extracted DNA, KAPA blood PCR kit Mix B (KAPA Biosystems, USA), 1330 and primers that target a ~605bp region of the replication associated (Rep) gene protein as previously described (Julian et al. 2012, Julian et al. 2013).
Faecal screening
Most chicks produced fresh faeces whilst being restrained or when placed in bags for weighing. These were collected into 1.5ml eppendorf tubes and refrigerated prior to 1335 processing (within 48 hours). Faecal samples were mixed with zinc sulphate solution
145 (specific gravity of 1.18) in a Fecalyzer pot (Vetoquinol, New Zealand). Further zinc sulphate was added until a meniscus formed, a coverslip was applied, and the sample allowed to stand for 20 min. Following this, the coverslip was applied to a glass slide and examined at 100x magnification using a standard microscope (Leica 1340
DM 1000, Leica Microsystems, Singapore).
Statistical analysis
All analyses were performed in “R studio” version 0.98.501.
Continuous data (clutch size, live young, fledge rate) were examined using the
Shapiro-Wilks test for normality. Differences of means for normal data were tested 1345 by the Welch t-test, and non-parametric data were tested using the Kruskal-Wallis ranked sum test to overcome issues of tied data.
When comparing parameters for nest type, wood was omitted as only one nest was found in a wooden box. The relationship between nest success (measured as ≥1 chick fledged), and the presence of nest mites or ≥1 parent with mange, was 1350 examined by the Odds Ratios (OR). Significance was measured by Fisher’s two-tailed exact test given sample sizes were <5 in some categories.
Haematological data was first examined for outliers, with one chick removed from analysis as a result of a high EWCC due to a panleukocytosis. All data was then examined using the Shapiro-Wilks test for normality. The effect of sex was 1355 investigated using the Welch t-test for differences of mean values in normally distributed data, and Kruskal-Wallis rank sum test with continuity for non-parametric data.
146 7.3 Results
Overall nesting success and nest types 1360
A total of 22 nests was observed during the period from finding the first nest on 20th
December 2012 to the last chick fledging on 21st February 2013. A total of 20 chicks fledged from all nest types, with category frequencies: plastic (n=6), wooden (n=1) and natural (n=15). Plastic nests were only occupied in the Wattle Valley Area (see
Figure 7.1). Natural nests were located across the island in a clay bank (n=1), at the 1365 base of flax (n=12), in a clump of Muehlenbeckia complexa (n=1), and in a fallen down pōhutukawa (Metrosideros excelsa) cavity (n=1).
Overall nesting success (one or more fledglings produced) was 45% (95%CI: 24-68%, n=10), with 33% (95%CI: 4-78%, n=2) success in artificial nest sites and 53% (95%CI:
27-79%, n=8) success in natural nests. The odds of a successful nest being natural 1370 versus plastic were not significant (OR: 1.87, 95%CI: 0.31-11.19, p=0.68). Overall reproductive parameters (clutch size, live young, fledge rate) for all nest types combined are presented in Table 1, with results of previous studies in RCP.
1375
147 Table 7.1: Clutch size, live young and fledge rate (± sd where available) for all nest types of Red-crowned Parakeets on Tiritiri Matangi Island 2012-13, compared with 1380 previous studies.
This study Greene Greene Ortiz- Ortiz- et al Cathedral Cathedral 2012-13 1989-90 et al 1990- 91 2004-5 2005-6
Nests 22 10 10 24 26
Clutch size 6.79 6.9 7.1 6.29 7.31
± 1.58 ± 0.33 ± 0.26
Live young 2.89 5.9 5.8 2.75 4.77
± 1.52 ± 0.49 ± 0.44
Fledge rate 0.91 2.0 2.6 1.04 3.27
± 1.11 ± 0.27 ± 0.41
Unsuccessful natural nests (n=7) were classed as “abandoned prior to hatch” (n=1),
“hen killed prior to hatch” (n=1), and “after hatching for unknown reasons” (suspect starvation or abandoned) (n=5). The hen killed prior to hatch was found moribund 1385 and ataxic underneath the nest with a penetrating wound to the back of the skull, and was subsequently euthanased. In the previous week, aggressive interactions had been observed between a Kingfisher (Halcyon sancta) fledging young adjacent to the
RCP nest, and the male RCP feeding the female on the nest.
Unsuccessful plastic nests (n=5) were “abandoned prior to hatch” (n=1), and 1390
“following hatch due to suspected starvation” (n=2), or “unknown” (n=2). Of the unknown complete brood losses in plastic nest boxes, both were found in disarray with dead chicks and eggs scattered or buried. Interspecific aggression had been
148 observed between a hihi (Notiomystis cincta) and the nesting RCP for two weeks prior to one nest failing, with a female hihi seen going in and out of the occupied 1395 nest box in the week prior to failure.
Reproductive parameters by nest type
Measures of reproductive output between plastic boxes and natural nest sites are presented in Table 7.2. As only one wooden nest box was found in use during the season, this was not included in statistical analyses. 1400
Table 7.2: Comparison of clutch size, live young and fledge rate between plastic nest boxes and natural nest sites for Red-crowned Parakeets nesting on Tiritiri Matangi
Island 2012-13.
Plastic Natural p-value
Mean sd n Mean sd n
Clutch size 6.67 1.97 6 6.88 1.36 8 0.74
Live young 2.5 2.1 6 3.08 1.26 13 0.84
Fledge rate 0.16 0.41 6 1.13 1.19 15 0.08
No significant differences were found between plastic and natural nest sites. 1405
Ectoparasitic mange, nest mites and reproductive success
An undescribed Dermanyssus species (Mesostigmata: Dermanyssidae) was detected in mounted samples from nest-mite positive nest boxes. Prevalence data for mange- affected parents and nest mite-positive nest boxes are presented in Table 7.3. Not all
149 parent birds were observed in sufficient detail to evaluate their feathers therefore 1410 not all could be classified as mange positive or negative.
Table 7.3: Prevalence data for infestation of Red-crowned Parakeet nests with nest mites, and ectoparasitic mange in parents observed on nests, on Tiritiri Matangi
Island 2012-13.
Factor Prevalence 95%CI n
Nest mites 64% 41-83% 22
Nest mites (plastic) 67% 22-96% 6
Nest mites (natural) 60% 32-84% 15
Mange (female) 50% 26-74% 18
Mange (male) 19% 4-46% 16
Mange positive nest (either parent with mange) 56% 31-79% 18
1415
The odds of being a female parent with mange, versus a male parent with mange, were not significant (OR=4.33, 95%CI: 0.9-20.6, p=0.08). When comparing natural nests versus plastic nest boxes with the outcome of nest mites, no significant difference was detected between the two nest types (OR (plastic) = 2.29, 95%CI:
0.32-16.51, p=0.64). 1420
A summary of findings for reproductive parameters between nest mite-positive and negative nests, and mange-positive and -negative parents (combined) are provided in Table 7.4.
150 Table 7.4: Comparison of reproductive parameters between nest mite-positive and - 1425 negative nests of Red-crowned Parakeets on Tiritiri Matangi Island 2012-13.
Clutch size Live young Fledge rate
Mean n Mean n Mean N
(+/- sd) (+/- sd) (+/- sd)
Nest mites 7 ± 1.4 10 3.3 ± 1.6 11 0.6 ± 0.9 14
No nest mites 6.3 ± 2.1 4 2 ± 1.7 7 1.4 ± 1.3 8
Mange* 7 ± 1.8 4 2.5 ± 2.1 6 1.4 ± 1.1 10
No mange 7.1 ± 1.5 7 3.25 ± 1.5 8 0.25 ± 0.7 8
* Where one or both parents had mange
No significant difference was found between nest mite-positive versus -negative nests for clutch size (p=0.54), live young (p=0.08), or fledge rates (p=0.15).
No significant difference was found between mange-positive versus mange-negative 1430 parents (where one or both parents were positive for moderate to marked signs of mange) when comparing clutch size (p=0.90) and live young (p=0.62). However, significantly higher fledge rates were observed in nests with mange-positive parents
(p=0.02).
Chick health results 1435
The weekly check frequency meant deceased chicks were often desiccated or maggot-infested so cause of death could not be determined. Those that were found freshly dead were in poor body condition, with empty crops.
In total, 17 fledging chicks were examined (from the 20 chicks fledged overall), comprising 10 females and 7 males, from 9 nests (n=3 from artificial nests, n=14 1440
151 from natural nests). All chicks were in normal condition with normal feathering for age. Faecal floats were performed on fresh faeces from 14 individuals, and no endoparasite (nematode) eggs or larvae were detected. Nine chicks came from nests with mites, with the remaining eight from nest mite-negative nests. Fourteen chicks fledged from nests where one or both parents had mange, however, none of these 1445 chicks showed clinical signs.
Haematology and nest mites
No significant effects were detected in the haemogram when comparing chicks from nests with mites versus those without for the EWCC (p=0.056), nor absolute counts of monocytes (p=0.21), basophils (p=0.33), lymphocytes (p=0.91) or eosinophils 1450
(p=0.45). However, chicks from nest mite-positive nests had a significantly higher mean absolute heterophil count (p=0.041).
Table 7.5 provides mean, median, sd, 10-90th percentiles and range data for nest mite positive versus negative nests, demonstrating the significantly increased absolute heterophil counts for chicks from nest mite positive nests. 1455
152 Table 7.5: Haematology of fledging Red-crowned Parakeet chicks from nest mite positive and nest mite negative nests on Tiritiri Matangi
Island 2012-13.
Mites No mites
Variable mean sd median 10-90 percentile range n mean sd median 10-90 percentile range n
EWCC x 109/L 7.5 1.67 7.8 5.8-9.5 4.8-9.7 9 5.6 1.04 5.0 5.0-6.4 5-6.8 3
Heterophil x 109/L 3.4 1.47 2.8 2.0-5.6 1.8-5.9 9 1.9 0.1 1.9 1.8-2.0 1.8-2.0 3
Lymphocyte x 109/L 3.2 1.23 2.9 2.2-4.9 1.6-5.1 9 3.3 1.13 2.7 2.6-4.2 2.6-4.6 3
Monocyte x 109/L 0.5 0.36 0.4 0.2-1.0 0.1-1.2 9 0.3 0.2 0.3 0.1-0.5 0.1-0.5 3
Eosinophil x 109/L 0.3 0.31 0.2 0.1-0.8 0.1-0.9 9 0.15 0.07 0.15 0.1-0.2 0.1-0.2 2
Basophil x 109/L 0.15 0.08 0.1 0.1-0.2 0.1-0.3 8 0.1 0 0.1 0.1-0.1 0.1-0.1 2
153 No significant effects were detected in the absolute counts for the haemogram of chicks when comparing sexes, including TWCC (p=0.54), monocytes (p=0.30), basophils (p=0.43), heterophils (p=0.62), lymphocytes (p=0.43) or eosinophils 1460
(p=0.57).
Analysis of haematology results for all chicks against reported values for adult RCP
(Chapter 3) revealed a significant difference for absolute heterophil counts
(p=0.004), which disappeared when chicks from nest mite-positive nests were removed from the dataset (p=0.53), and increased in significance where only chicks 1465 from nest mite-positive nests were analysed against adult RCP (p=0.002).
BFDV results
No nests were found to be BFDV-positive based on testing of adult feathers from the nest (n=19), although 2 nests were not tested, as feathers could not be obtained. All fledging chicks tested negative by PCR of blood and feather for BFDV infection 1470
(n=17). No chicks were observed with classical signs of BFDV infection. Given a design prevalence for BFDV in the population of 1-4% (based on findings in Chapter
4), and a test sensitivity of 0.95%, the probability of detecting BFDV with a sample size of 19 nests with 95% confidence would be 0.42-0.52 (Sergeant 2015), where the unit of interest is the nest. 1475
154 7.4 Discussion
Overall nesting success 1480
This study aimed to assess the impact of specific parasites on nesting success, however unusual climatic conditions during the study season (2012-13) are likely to have affected the results. Severe dry climatic conditions were reported for the
Auckland region in summer of 2012-13, with extreme soil moisture deficits and the
3rd lowest total summer rainfall recorded at the closest weather station to Tiritiri 1485
Matangi Island (NIWA 2013). Nesting success in this study, as measured by clutch size, live young and fledge rates, was similar to that reported for RCP in the 2004-5 nesting season on Tiritiri Matangi Island (Ortiz-Catedral and Brunton 2008).
However, our study found significantly lower numbers of live young (p<0.0001) and fledge rates (p<0.0001) when compared to the more productive 2005-6 nesting 1490 season by the same authors, providing further evidence of marked inter-annual variation in nesting success in RCP. Breeding success is known to be affected by climate in other parrot species, likely mediated by impacts on resource availability, with higher rainfall improving breeding success in the Crimson Rosella (Platycercus elegans) (Krebs 1998). Breeding success overall in our study appeared to be at the 1495 low end of reported data for RCP in New Zealand (Table 7.1), most likely attributable to the extreme and adverse environmental conditions.
Intraspecific and conspecific egg destruction has been reported in parrots, including the Eastern Rosella (Krebs 1998). We did not observe significant competition between nesting parakeets and, in fact, successful nests were observed within 1500 approximately 20m of each other. However, one nest failed when the female was
155 found moribund adjacent to a Kingfisher nest, with the single penetrating wound to the skull suggestive of a Kingfisher attack. Aggressive interactions were observed between the two species the week preceding this event, and kingfishers are known to be highly territorial with recorded attacks on intruder model birds (Legge et al. 1505
2004). Interspecific aggression was observed between a hihi pair and nesting RCP, with failure of that nest preceded by two weeks of observed interactions including the hihi entering the occupied nest box. Whilst interspecific aggression cannot be confirmed as a cause of nest failure in RCP, these observations warrant further investigation and serve as a reminder of the wide range of potential causes of nest 1510 failure where multiple species compete for nesting sites.
Nest mites
Impact on breeding success and chick health
An undescribed Dermanyssus species was detected in 64% of all nests (95%CI: 41-
83%), with no significance detected between prevalence of nest mites in plastic 1515 boxes versus natural nest sites. Although these results suggest nest mites are as likely to be found in nest boxes as wild nests, sample sizes were small which may lead to type II errors (e.g., failure to detect an effect) (O'Brien et al. 2009). Presence of dermanyssid nest mites did not affect clutch size, number of live young or fledge rates of RCP in this study, similar to results found for saddleback infested with both 1520
O. bursa and an unidentified Dermanyssus sp. on Tiritiri Matangi Island (Stamp et al.
2002). However, chicks fledging from nest mite-positive nests had significantly higher absolute heterophil counts in comparison with reference ranges for adult RCP
(Chapter 3) and chicks from nests without nest mites. Heterophils may be stimulated
156 by bacterial, viral or parasitic agents during the acute inflammatory response in birds 1525
(Mitchell and Johns 2008), and so this study documents a plausible host inflammatory response correlated with ectoparasitism in RCP fledglings from nest mite positive nests. The short and long term fitness costs of this host response are unknown.
Reports of the impact of nest mites in other species, both locally and internationally, 1530 have been variable. Ornithonyssus bursa affects survival, growth rates and asymptotic mass (maximal growth) in hihi (Ewen et al. 2009), fledge dates and size in
North Island Robins (Berggren 2005), and leads to delayed reproduction, impacts on nestling size or body condition, nestling mortalities, and increased reproductive costs in Barn Swallows (Møller 1990). This latter study found adult birds were reluctant to 1535 re-nest in nest sites previously infested with mites. We did not examine asymptotic mass, fledge dates or growth rates specifically and so cannot infer likely impacts on them, however survival and condition were not influenced by nest mite presence.
Costs of parasitism are often cryptic and may be synergistic with weather or resource availability (Stamp et al. 2002), as evidenced by compensatory effects of 1540 dietary carotenoids (reflecting nutritional quality) on the cost of ectoparasitism in hihi (Ewen et al. 2009). Differences between studies of nest mites in wild birds may be related to the life cycle and feeding strategies of O. bursa, which feeds both day and night, compared to dermanyssids that feed at night for short periods. It may also reflect host species differences in tolerance or immune response to the parasite. 1545
157 Biology and epidemiology of nest mite infestations
The epidemiology of D. gallinae infestations, coupled with knowledge of the biology of RCP, provide insight into factors that are likely to play a role in the impact and 1550 ecology of the undescribed Dermanyssus sp. found in RCP nests in this study, including source and transmission-factors that may lend themselves to targeted conservation management actions. Dermanyssid mites are haematophagous ectoparasites of birds (Roy and Chauve 2010), with similar behavioural and morphological characteristics to Ornithonyssus spp. (Mesostigmata: Macronyssidae) 1555
(Di Palma et al. 2012). The life cycle and epidemiology of the Dermanyssus sp. found in RCP nests are currently unknown, however they are likely to be similar to those of the well-described D. gallinae.
Dermanyssus gallinae spends the majority of its life cycle off the bird host, favouring cracks, crevices and nest boxes or feeding areas where birds congregate (Chauve 1560
1998). Blood feeding is mainly by nymphs and females, rather than males, and lasts for 1-1.5 hours per night (Chauve 1998, Meyer-K hling et al. 2007). Females lay approximately 30 eggs over several clutches in their lifetime, and in warm conditions the complete lifecycle (egg-egg) takes 7 days (Chauve 1998), so numbers can increase rapidly (Meyer-K hling et al. 2007). Mites can survive up to eight months 1565 off the host (Chauve 1998). Wild birds, and equipment such as nest boxes may enable indirect transfer of mites and act as reservoirs for re-infestation of an environment (Meyer-K hling et al. 2007). Reported clinical impacts in poultry due to
D. gallinae infestation appear to be related to duration of exposure, ranging from irritation and anaemia, decreased laying, or deaths in breeding birds housed for long 1570
158 periods (Meyer-K hling et al. 2007), through to no observable impacts in broiler birds housed for only 52-55 days (Marangi et al. 2012).
RCP pairs are reported to use regular roost sites during the pre-breeding season, with these roosts sometimes used as nests by other bird species (Greene 2003). Thus roost sites may form an important source or over-wintering location for nest mites, 1575 as well as providing a transmission point for other diseases or parasites identified at the study site, including BFDV and mange mites (Chapter 4 and 5). The artificial nest boxes used in this study were free of mites prior to occupation by RCP, and mites were not detected during the egg phase in this study or that reported by (Stamp et al. 2002), suggesting mites are brought into the nest by adult birds. As noted by 1580
Stamp et al. (2002), mite numbers and therefore impacts are likely to be higher on relatively less-feathered chicks who are unable to preen and remove ectoparasites as efficiently as adults.
Treatment and control of nest mites
Treatment and control options for nest mites need to be evaluated in the context of 1585 the impact of the parasite on the host species, conservation status and management objectives for the host species, efficacy and toxicity of the control method chosen, and the relative cost (financially and at the population scale) of doing nothing.
Targeted studies are warranted in species of conservation significance, to determine the need for treatment or control. Acaricides are the primary method for control of 1590 nest mites in poultry (Marangi et al. 2012). Over 35 different agents have been described for control of D. gallinae, including environmental or individual bird treatments, with notable food safety and residue issues (Chauve 1998) leading to
159 widespread bans on the use of many of these compounds (Chirico and Tauson 2002,
Marangi et al. 2012). Acaricides are generally designed to be used between 1595 reproductive cycles when birds are not present (Marangi et al. 2012), and are not licensed for use in wild birds. Resistance to commonly used acaricides is widely reported (Chauve 1998), as are issues of bioaccumulation and impacts on reproductive success. For example fipronil, a common veterinary ectoparasiticide currently used topically on nestling hihi to control O. bursa (Ewen et al. 2009), was 1600 associated with significantly reduced hatching success in zebra finches when used at sub-lethal doses of 1mg/kg (Kitulagodage et al. 2011).
Therefore, whilst mite control options exist, they should be carefully evaluated prior to use for toxicity in all species that may be exposed, as well as efficacy to be sure their use is justified and to reduce development of resistance (Chirico and Tauson 1605
2002). For species such as the critically endangered Malherbe’s Parakeet, maximising reproductive potential in translocated populations where nest boxes are provided may warrant management of nest mites based on the precautionary principle.
Reducing mite loads between seasons via good nest box design and biosecurity (e.g. thorough cleaning and disinfection) should remain a first point of control. Given nest 1610 mites appear to be widespread in the environment globally, and on Tiritiri Matangi
Island specifically, eradication is not a feasible (nor necessarily indicated) option for management. The preference of mites for hidden cracks provides a useful option for control, and research into the use of acaricide impregnated strips (Chirico and
Tauson 2002) in door frames or hidden compartments of nest boxes would be a 1615 useful study for control options that limit bird exposure to these potentially harmful
160 agents. A 95% reduction in D. gallinae mite numbers was recorded using 2% metriphonate impregnated strips replaced weekly during an 8 week trial at a poultry farm (Chirico and Tauson 2002). Given dermanyssids are only on the host for short periods to feed, it seems more logical to attempt treatment where the mites 1620 congregate, as treating the host results in poor exposure of mites and thus low treatment effect (Chirico and Tauson 2002), as well as secondary toxicity issues.
Nest mites and host resistance
One of the few long term studies to evaluate host-parasite co-evolution in nest mites in wild birds, found evidence for micro-evolutionary changes with development of 1625 host-parasite resistance linked to a reduction in abundance and prevalence of the nest mite O. bursa in Barn Swallows (Møller 2002). Consequently the reproductive costs of parasitism changed significantly during the study, with strong negative effects detected in 1988 not repeated in 1999 (Møller 2002). Resistance to nest mites may be mediated by the host immune response leading to mite death, or 1630 physical or biochemical changes in the skin that prevent blood feeding as hypothesized in Moller (2002) and demonstrated in poultry infested with O. sylviarum (Owen et al. 2009). A significant 50.6% increase in mite mortality has been reported in an in vitro study on the immunisation of poultry against D. gallinae
(Harrington et al. 2010), although experimental evidence of in vivo host immunity 1635 appears to be lacking for this mite compared with other species (Sparagano et al.
2014). It is unknown if RCP may develop host immunity to the Dermanyssus sp. found that may reduce nest mite presence and abundance in older birds nesting.
However it raises an interesting question as to the potential for population scale
161 resistance to the parasite, as well as maternal transfer of antibodies that may 1640 protect chicks. If the Dermanyssus sp. found is a recent introduction, it is possible that co-evolutionary processes may mediate the impacts of this parasite over time, without the need for intervention. Long term studies are required to accurately infer these dynamics, and it is important to note that relatively high generational turnover of mites creates the potential for sudden increases in mite abundance if a more 1645 virulent form emerges (Møller 2002). Future studies could investigate mite prevalence and abundance, as well as adult and chick impacts, in nests of RCP with and without previous exposure to the mite.
BFDV and nesting success
BFDV, a well-described conservation threat for wild parrots with highest impact in 1650 the juvenile age class (Raidal 1995), was not detected in feathers lining the nest nor in any of the fledging chicks in this study. BFDV has only been detected at a low prevalence (1-4%) in RCP on Tiritiri Matangi Island over multiple years (Chapter 4), and therefore samples sizes in this study were too low to detect a BFDV positive nest at this design prevalence. The effects of BFDV in wild RCP chicks and consequences 1655 for reproductive parameters therefore remain unknown, however this study provides further evidence that this virus is not currently at a prevalence that impacts a large number of nests in the study site. The cause of death in chicks found dead in nests was difficult to assess in most cases as the frequency of checks (weekly) meant most chicks were in advanced states of autolysis or decay by the time of discovery. In 1660 the absence of histopathological examination infectious causes that may lead to failure to feed and therefore clinical signs at the time of death that mimic starvation
162 cannot be ruled out. However, it is likely that starvation was the cause of death in most instances, given the resource constraints and climatic conditions, as well as commonly finding chicks with empty crops and weak or in poor condition prior to 1665 death. Chicks that survived to fledging were in good body and feather condition, with no evidence of significant endoparasitism on faecal screening.
Ectoparasitic mange, breeding success and chick health
In 45% (95%CI: 21.5-69.2%) of nests observed, one or more parent birds had ectoparasitic mange based on the case definition. Proportionally but not significantly 1670 more females than males were affected, although we note the potential impact of small sample sizes (O'Brien et al. 2009). Either parent having mange did not negatively affect reproductive parameters in this study. The life cycle of the species of mite associated with ectoparasitic mange in RCP on Tiritiri Matangi Island (P. janssensi) is unknown, but it is likely to be similar to other knemidokoptinids (Fain 1675 and Elsen 1967). Mites in this subfamily spend their entire lifecycle on the host, and require close contact for transmission (Wade 2006) making the nest environment and pair bonding the most likely point for transmission between individuals, although water and feed areas where birds congregate may be involved to a lesser extent. Host factors such as nutritional state and physiologic stress, including 1680 reproduction, may influence the expression of mange. Therefore, it is not surprising that clinical signs of feather loss appeared to subjectively worsen in some individuals during the breeding period. Knemidokoptinid mange, unlike other parasitic diseases, is more prevalent in adult birds, suggesting latency from mites acquired as a juvenile bird to the development of clinical signs (Jaensch et al. 2003). This study did not 1685
163 detect any clinical signs of mange in 14 chicks fledged from nests with one or more parents affected by moderate to severe ectoparasitic mange. However, as sub- clinical infestations have been commonly found in adults (Chapter 5), it is likely they become asymptomatic carriers of the mite due to close contact with infested parents. Future studies could incorporate a skin biopsy to determine host status with 1690 respect to skin mites at the time of fledging, and follow up monitoring to investigate likelihood of developing mange as well as impacts on survival.
Nest types
Reproductive parameters did not differ significantly between natural nests and plastic nest boxes. Nest sites did not appear to be limiting in this study as only 16% 1695 of available plastic nest boxes were used, and these were only occupied in an area where other cavity options were limited. We suspect where natural choices are available birds will preferentially choose these. The wide variety of natural nest types used on Tiritiri Matangi Island in this and previous studies (Greene 2003, Ortiz-
Catedral and Brunton 2010a), demonstrates the flexibility of RCP nesting choices 1700 based on available habitat. This study provides evidence that RCP will use artificial nest boxes within months of their placement but only where natural sites were substantially reduced or lacking. Nest box design is critical in other species on the island such as the saddleback (Stamp et al. 2002), whereas RCP have now been shown in this and other studies to use a variety of nest box designs (Ortiz-Catedral 1705 and Brunton 2010a). Plastic nest boxes provide an advantage over wooden designs with respect to their durability, and their non-porous nature that enhances cleaning and disinfection. Future studies should focus on collection of temperature and
164 humidity data comparative to natural nests, to ensure the design does not introduce microclimate impacts on parent birds and reproductive success (Lei et al. 2014). 1710
7.5 Conclusions
This is the first targeted study looking at specific disease and parasite constraints on
RCP breeding success. Ectoparasitic mange in parent birds did not correlate with decreased breeding success, nor did chicks fledging from mange positive nests have clinical signs of mange. A Dermanyssus species of nest mite was identified in both 1715 natural and artificial RCP nests, with the first report on prevalence across these nest types in the wild. There was a correlation between nest mites and heterophilia in
RCP chicks, although no negative impacts were found for clutch size, live young or fledging success where nest mites were present. Further work is recommended to determine any impacts on fledge date, growth rates, and anaemia due to 1720 haematophagy as measured by packed cell volume. The potential role of nest mites in transmission of disease on the island, as well as public health implications for researchers and rangers should also be noted. It is likely that nest mites, including O. bursa and the undescribed Dermanyssus sp., are ubiquitous in the environment on
Tiritiri Matangi Island, and potentially across New Zealand. Further work to 1725 characterise nest mites by location and host species in NZ would aid in decision- making as to the biosecurity risks of moving birds that may be carrying these mites.
For reintroduction or translocation programs, treatment for nest mites at the time of movement may be recommended if destination sites are considered free of this parasite. As discussed, there are multiple considerations needed prior to instigating 1730 control and treatment options for nest mites and other parasites or diseases. Tiritiri
165 Matangi Island provides an option for a globally relevant multispecies study of nest mite impacts, especially given there is long term population and/or nesting data for many of the known infested species including the North Island Robin (Petroica longipes), hihi, Saddleback (Philesturnus carunculatus) and now RCP. Future studies 1735 of breeding success in RCP and other parrot species should incorporate health assessment, including parasitism, to ensure that all impacts on reproductive success have been carefully evaluated.
1740
166
1745
CHAPTER 8:
GENERAL CONCLUSIONS
1750
167 8.1 Health and disease in conservation management
This study draws attention to the ongoing need to include health and disease monitoring for all parakeet species to determine the cause of observed clinical syndromes, to provide baseline data, and to detect presence or changes of endemic levels of target pathogens. This is particularly important for endangered species such 1755 as Malherbe’s Parakeet (Orange-fronted Parakeets, OFP, Cyanoramphus malherbi) and Forbes Parakeet (Cyanoramphus forbesi), given the increased risk of extinction by disease in these threat classes (Heard et al. 2013). Through incorporation of health and disease surveillance into wider recovery programs, disease can be appraised against other threats, ideally in combination with population data, which 1760 will help inform the threat levels and prioritise actions. Performing a Disease Risk
Analysis (DRA) for all threatened species programs will identify knowledge gaps to drive research objectives and improve certainty around high priority threats (Jakob-
Hoff et al. 2014). Threatened species programs rely on these sorts of adaptive management frameworks that update states of knowledge based on focused 1765 research, to improve conservation outcomes (Armstrong et al. 2007).
A lack of evidence is not an acceptable reason to exclude disease as a threatening process, particularly where limited or no disease screening has taken place. Like all threats, evaluation of disease risks should be evidence-based where possible, and thereafter expert opinion sought. It is clear that for many species, including New 1770
Zealand parakeets, predation by introduced species and habitat alterations are still primary threats (Innes et al. 2010) that must be addressed. However, there should be efforts to investigate the significance of disease in any threatened population,
168 particularly given the indirect and sub-lethal effects that disease may have on populations including reduced reproduction (Spalding and Forrester 1993), and 1775 increased risk of predation as seen in BFDV infection where flight feathers are affected (Collings et al. 2015). BFDV is also more likely to have a high mortality rate in juvenile or nestling parrots (Raidal 1995, Schoemaker et al. 2000, Jackson et al.
2014a, Collings et al. 2015). Significant losses in this group will not be detected unless specific monitoring of nest sites and nesting success takes place. Also, without 1780 appropriate screening for disease the death of nestlings may be falsely attributed to starvation or other causes.
The difficulties of detecting wildlife carcasses during mass-mortalities due to disease has been experimentally reviewed, with evidence that even 1-day after a simulated event, mortalities are likely to be seriously underestimated (Wobeser and Wobeser 1785
1992). Depending on the disease syndrome, location of the species of interest, and level of population monitoring, significant disease-induced mortality events have the potential to go undetected (Ewen et al. 2007). The scale of population monitoring and longevity of datasets required to effectively identify such events has been highlighted by examples such as the BFDV emergence in the Mauritius Parakeet 1790
(Psittacula eques) (Kundu et al. 2012), and the 2006 outbreak of Salmonella
Typhimurium DT195 in hihi (Notiomystis cincta) on Tiritiri Matangi Island (Ewen et al.
2007).
A barrier to inclusion of disease surveillance, particularly for translocations and reintroductions, is the cost of disease sampling and diagnostics (Makan 2009). Cost- 1795 benefit analyses, which are regularly performed for livestock diseases where
169 economic benefits of screening and interventions are more readily quantified
(Berentsen et al. 1992), are rarely performed for wildlife diseases due to a lack of data on the impacts of disease or management interventions (Karesh 1993), let alone the ability to assign value to various outcomes. Where these analyses do 1800 involve wildlife it is usually in the context of threats to human and livestock health, with benefits to wildlife not represented in the analyses thereby masking the true benefits (Mwacalimba et al. 2013).
However, one of the potential costs of failing to effectively mitigate disease risks in wildlife movements will be the intrinsic financial cost for delivery of the 1805 translocation/reintroduction program, as well as the potential failure of the program itself. The case of the Californian condor (Gymnogyps californianus) reintroduction was a sobering reminder of the financial impact of disease, where environmental lead poisoning severely impeded recovery efforts including failure to meet program objectives of a self-sustaining wild population (Cade 2007). Whether disease risks 1810 should be included in plans for translocations and reintroductions should no longer be an issue for debate. How disease is incorporated and what value is assigned to it
(both financial and based on objectives of key stakeholders), is a different matter, and should be an iterative process involving tools for decision-making and expert opinion. Effective disease surveillance should incorporate strategies for the outcome 1815 of any disease screening at the planning stage, including how positive data will affect the movement, as well as risk communication to stakeholders (Thrusfield 2007).
Another benefit of health and disease screening for translocations and reintroductions is that the increasing baseline data derived across regions and
170 species will better equip conservation managers to interpret findings from future 1820 screening, and to make decisions as to risks for source and recipient populations in
Disease Risk Analyses. If used wisely, screening data will refine future activities based on real data rather than speculation. In the case of the RCP, as a result of this and previous studies, we are now in a position to make recommendations and decisions with a higher degree of certainty for this species in the Hauraki Gulf region for 1825 multiple pathogens, and across New Zealand for BFDV, a globally recognised conservation threat for wild parrots (Borthwick 2005, Kundu et al. 2012, Peters et al.
2014).
8.2 Updated knowledge of disease in wild Red-crowned Parakeets
The progressive thinking of previous conservation managers, researchers, and 1830 wildlife veterinarians in New Zealand provided a baseline of information for specific pathogens of RCP (Jakob-Hoff 2002, Ortiz-Catedral et al. 2009a, Ortiz-Catedral et al.
2011), particularly Beak and feather disease virus (Ha et al. 2009, Ortiz-Catedral et al.
2010, Massaro et al. 2012). Results of this study have updated our knowledge on diseases of parakeets, including the description of a skin mite (Procnemidocoptes 1835 janssensi), previously unreported in New Zealand, causing epizootic mange. We have also expanded our understanding of existing diseases of concern such as Beak and
Feather Disease, with a suggested downgrading of the threat status of the associated virus (BFDV), whilst acknowledging the risk this virus will continue to pose to small, isolated and naïve populations (Jackson et al. 2014b), as well as captive breeding 1840 programs. It is important, however, to remain vigilant however for mutation or recombination events that lead to enhanced virulence, a means by which BFDV may
171 become a more serious threat in the future (Julian et al. 2013, Sarker et al. 2014c).
Although our preliminary results did not find a significant impact of nest mites nor mange positive parents on fledging success, we recommend further research into 1845 the haematological impacts of nest mites on fledging chicks, specifically Packed Cell
Volume and Total Protein measures. Further research with larger sample sizes over successive years are needed to provide sufficient confidence regarding the potential impacts of mange and nest mites on nesting success in RCP, and to resolve for confounding environmental factors as well as underlying inter-annual variation in 1850 breeding success.
Considering non-infectious diseases of RCP, the creatine kinase (CK) findings from this study warrant further research to determine if these levels correspond with an increased susceptibility to capture myopathy during handling events in this species.
We note that normal CK levels can vary substantially between avian species, and that 1855 without focused studies in the species of interest, it is difficult to correlate absolute values with risk of capture myopathy. Given our findings that CK continues to rise in response to handling, it would be valuable to compare CK levels in birds sampled under manual restraint, versus those sampled under anaesthesia, to determine the association between each method and rising CK levels. Targeted observations of 1860 banding combinations or recapture of birds will enable correlation between CK levels and survival post-sample collection. Repeat observations must use appropriate time frames given the acute and chronic forms of capture myopathy (Williams and Thorne
1996a). The territoriality and ease of observation for RCP on Tiritiri Matangi Island would make such a study feasible. The findings reinforce the importance of 1865
172 minimising handling times when sampling wildlife, using appropriate handling technique, keeping noise and stress to a minimum, and conducting studies during the cooler part of the day (Williams and Thorne 1996a).
This study has highlighted the dynamic state of disease in wild RCP populations.
Substantial inter-annual variation was observed in prevalence of feather loss due to 1870 ectoparasitic mange, as well as a low endemic prevalence of BFDV in the population of RCP on Hauturu-o-Toi/Little Barrier Island (LBI) where previous point prevalences have been significantly higher (Ortiz-Catedral et al. 2009b). This serves as a reminder that a single estimate of disease in a population may fail to detect disease, or to accurately describe disease trends, and this compromises our ability to correlate 1875 disease expression with other factors including anthropogenic and environmental drivers. It also compromises or restricts the validity of decisions made by conservation managers on the basis of very limited disease surveillance.
Although some may view diseases of wildlife as having limited relevance to broader ecosystem health and function, it is now broadly accepted that wildlife population 1880 health and ecosystems are intrinsically linked, with anthropogenic influences on climate, habitat and resources increasingly recognised as environmental drivers of disease emergence (Daszak et al. 2001). It is therefore important to consider disease expression in populations as a potential sentinel for broader ecosystem changes
(Jessup et al. 2001, Jones et al. 2009). It is most common to monitor only diseases 1885 that have a perceived impact on population viability. Whilst this makes inherent sense, it overlooks the opportunity to track diseases, such as mange, that may be sub-lethal but are likely to be environmentally driven, thereby providing information
173 on wider ecosystem structure and function such as resource availability and habitat quality. Disease surveillance should not be legitimised only where significant 1890 population impacts occur. Rather, in some cases, it has the potential to be used as another tool by which to infer overall population health and resilience, as well as responses to ongoing climate variability.
8.3 Methods for wildlife disease surveillance
Importantly, sample sizes in this study were sufficient to determine presence or 1895 absence of disease down to a relatively low predicted prevalence (target of 5-10% per sampling session), as well as provide data on disease trends over different seasons. Repeated cross-sectional studies are more useful than single point estimates for detecting disease trends and informing conservation managers. It is critical that disease surveillance follows appropriate epidemiological planning to 1900 ensure sample sizes are adequate or, failing that, that interpretation of negative results highlights the likelihood of disease detection given not only the predicted or prior prevalence of the disease of interest, but also the test characteristics used to screen for the disease (Nusser et al. 2008, Ryser-Degiorgis 2013). Appropriate sample selection and storage are also fundamental to disease surveillance, and may 1905 require consultation with disease experts, diagnosticians and diagnostic laboratories.
Screening a blood sample alone for an observed skin disorder would not have been sufficient to detect the cause of feather loss in this study, neither would a skin scraping have necessarily detected the parasite. Further, the pathology observed in the skin biopsy was vital to enabling evaluation for causal links between pathogens 1910 found, and the extent, mechanisms and nature of related changes (Ryser-Degiorgis
174 2013). We recommend obtaining appropriate samples for histopathology to augment interpretation of diagnostic screening whenever possible.
Obtaining the right sample in this study to diagnose the cause of feather loss ethically and rapidly would not have been possible without the use of field 1915 anaesthesia. Field general anaesthesia proved a safe and reliable method for sampling RCP, taking an average of 12 minutes from induction to recovery. During this time a sample of blood, feathers and faeces were obtained; a full physical examination performed including auscultation of the air sacs and examination of the oral cavity and eyes; temperature, heart rate, respiratory rate and morphometric 1920 measurements obtained; a skin biopsy taken with suturing of the remaining incision; subcutaneous fluids given; and a series of four standard photographs taken of the bird. To do all of this on a conscious bird would result in significant handling stress to the individual, and the inability to undertake more invasive sampling such as a skin biopsy. We recognise the logistical limitations of field anaesthesia, as well as the 1925 need for an attending veterinarian, therefore it may not be possible in all cases.
However we recommend the use of field anaesthesia where appropriate to reduce risks of capture myopathy in at-risk species, to facilitate more invasive procedures, and most importantly, to reduce stress in the individual (West et al. 2007).
Access to passive or opportunistically obtained samples to augment targeted 1930 sampling remains a key part of disease surveillance in wildlife for a number of reasons (Ryser-Degiorgis 2013). Primarily, there are ethical and welfare reasons to either piggy-back or bank samples from existing research for analyses (including retrospective), as this adheres to the principals of reduction and refinement set out
175 in the Code for the Care and Use of Animals for Scientific Purposes (NHMRC 2013). 1935
Across New Zealand there are a great number of research studies being conducted on native and exotic wildlife that include obtaining blood and other samples appropriate for disease screening (and other objectives). Whilst it may be ambitious to centralise a bank of biological samples, and notwithstanding issues of ownership and intellectual property, if appropriately stored and curated, these samples could 1940 be accessed for disease surveillance as well as other future research objectives e.g. population genetics. Importantly, banked samples provide an opportunity to retrospectively derive baseline data, which can prove useful in determining if newly detected diseases are truly novel (Ryser-Degiorgis 2013). Further, analyses
(particularly using molecular tools) of stored material can help infer when, where 1945 and how a pathogen may have entered the wild. The retrospective detection of chytrid fungus (Batrachochytrium dendrobatidis) in archived samples from captive mortalities in the Mallorcan midwife toad (Alytes muletensis) breed for release program, demonstrated the source of this pathogen for the wild population (Walker et al. 2008). We therefore urge researchers, conservation managers, and those 1950 working with wildlife species, to seek advice on appropriate sample sizes when collecting biological specimens, such as feathers, blood and faeces, that may be stored for future use and provide a retrospective biological sample bank for RCP and related parakeet species in New Zealand.
Wildlife rehabilitation centres, biodiversity rangers, community conservation groups, 1955 the general public, and veterinary centres that accept wildlife, all form a frontline for wildlife disease surveillance, as this group represents the people on the ground
176 observing wildlife and potentially identifying disease processes or unusual mortality events through passive surveillance (Nusser et al. 2008). Passive surveillance will be enhanced through targeted educational campaigns to improve sensitivity to disease 1960 detection, specifically what is ‘abnormal’, and when and who to contact if a disease event is suspected. This study benefited from observations and samples from the
Supporters of Tiritiri Matangi Island, DOC rangers, Whitford Wildbird Care, Auckland
Regional Council rangers, ZEALANDI-Karori Sanctuary, and other interested members of the public. In particular, screening of stored blood and feather samples from 1965
Karori Sanctuary and DOC provided not only robust measures of disease prevalence at those sites, but also sequence data that allowed phylogenetic analysis of the flow of viral strains within and across regions. Whilst passive surveillance is widely utilised for collecting data on wildlife diseases (Nusser et al. 2008), including as sentinels for emerging infectious diseases affecting humans and wildlife (Eidson et al. 2001a, 1970
Eidson et al. 2001b), it is important to understand the reporting bias and limitations of using this sort of data (Thrusfield 2007). Active or targeted surveillance should be instigated when more accurate information on a disease is required such as true prevalence and risk factor analysis (Thrusfield 2007, Ryser-Degiorgis 2013).
8.4 Biosecurity and public health for wildlife research 1975
Biosecurity, based on a comprehensive understanding of the risks, is vital to ensure research activities do not become a means by which pathogens can artificially spread within a population (Cornish et al. 2001), in new host species or geographic ranges
(Walker et al. 2008), and do not facilitate mixing of pathogen or parasite strains.
Unlike other biosecurity risks such as seeds and pest animals, the majority of 1980
177 pathogens and parasites are small if not microscopic, making detection of these agents on research materials and equipment more difficult. Evidence suggests humans may play a role as fomites in pathogen spread for major biodiversity diseases such as white-nose syndrome in bats of North America (Warnecke et al.
2012), a reminder of how critical it is to prevent research equipment and personnel 1985 becoming vectors for pathogen transfer. This means developing a comprehensive biosecurity plan prior to research activities, that uses dedicated equipment where possible, and disinfection with effective biocidal agents to protect against known and unknown hardy pathogens, as demonstrated in the methods for this study. Such a biosecurity plan should include wearing gloves or washing hands and equipment 1990 between individuals and sites, to prevent cross-infection and sample contamination that may affect accuracy of results reported. Biosecurity plans must also include broader issues of spreading pest plants and animals, however in the author’s experience these biosecurity issues are usually more widely acknowledged and incorporated into research activities. Better research biosecurity will be aided by the 1995 inclusion of mandatory risk assessments and risk mitigation strategies for pathogen transfer in research permit applications.
Humans are not immune to risk when it comes to wildlife diseases, with 71.8% of emerging zoonotic diseases coming from a wildlife source (Jones et al. 2008).
Although not a cause of significant human disease, this study identified a nest mite 2000 species from a genus known to cause transient dermatitis in exposed humans (Auger et al. 1979), as well as having the ability to transmit infectious agents such as
Salmonella sp. and viruses between poultry (Valiente Moro et al. 2009). Birds in New
178 Zealand can harbour a range of pathogens that are able to infect humans, including
Chlamydia psittaci (Gartrell et al. 2013), Salmonella sp. (Alley et al. 2002, Ewen et al. 2005
2007), and Campylobacter sp. (French et al. 2009), with potentially serious disease consequences, particularly in the immune-compromised. Therefore the health risks to people must also form a part of risk assessments when working with wildlife.
These should include communication of those risks to personnel in the field or in facilities working with wildlife, and steps to prevent zoonotic disease transfer, such 2010 as wearing gloves or washing hands after handling wildlife or equipment that may have been contaminated by faeces, blood or other infectious material.
New Zealand must continue to mitigate risks of disease incursion through the captive parrot trade (legal and illegal) in New Zealand (Gartrell et al. 2003), with passive and targeted surveillance of captive and wild parrots, and capacity building from a 2015 diagnostic perspective. It is most likely that an exotic disease will be detected in captivity prior to the wild, therefore a risk-based surveillance design (Stark et al.
2006), that targets higher risk locations or facilities, or birds entering veterinary care with specific disease criteria, would improve sensitivity of disease detection. An informed conservation community including rangers, researchers, managers working 2020 with species recovery, aviculturalists, general public, wildlife rehabilitators and veterinarians will ensure risks and signs of key parrot diseases are recognised and responded to in a timely manner. However in New Zealand there is a need to improve diagnostic capacity for parrot diseases (Jakob-Hoff 2002), with commercial molecular tests only currently available for avian malaria, BFDV, and Chlamydia 2025 psittaci. Although suspicious pathology is likely to be detected and followed up by
179 avian pathologists in New Zealand (Gartrell et al. 2003), the availability of validated commercial screening for the full range of parrot viruses (e.g., bornaviruses, polyomaviruses, herpesviruses) reported in captive birds would assist in surveillance and early detection. The necessary response to an incursion, including movement 2030 restrictions and control methods, has been outlined previously (Gartrell et al. 2003), and depends on the nature and source of the pathogen discovered.
8.5 Research recommendations
Future research should focus on linking ecological and environmental drivers of disease expression in RCP populations, especially given apparent fluctuating 2035 population sizes driven by breeding events that may be governed by resource availability (Elliot et al. 1996). Population density and dynamics are drivers of disease transmission and expression (Aiello et al. 2014). Correlating these population factors and environmental variables - including resource availability and climate - with disease outcomes over a longer timeframe will provide greater insight into disease 2040 ecology and impacts in RCP.
Given the expansion of the Eastern Rosella (Platycercus eximius) on the North Island, this and other sympatric parrot species should be included in future surveillance, as they may form an epidemiologically important reservoir species for pathogens that would otherwise be subject to endemic or epidemic fade out in a small population 2045
(Lloyd-Smith et al. 2005). Sympatric species can also provide an initial surveillance tool where there are concerns over capture impacts on highly threatened species.
For example, Yellow-crowned Parakeets (Cyanormpahus auriceps) that share habitat with the remaining populations of critically endangered OFPs could be sampled for
180 targeted pathogens (e.g. BFDV and skin mites), as a preliminary investigation into 2050 circulating diseases of interest in that region. Results from sampling could be used to help focus research for OFPs and thereby reduce impacts on this species. However when using other species as sentinels it is critical to think about the biology and ecology of the host and epidemiology of the pathogen of interest. For example, to date BFDV has not been found in kākā (Nestor meridionalis)(Massaro et al. 2012), 2055 suggesting they are either highly susceptible to the virus or inherently resistant.
Either way, surveillance in this species may not reflect circulating virus as effectively as surveillance in the Eastern Rosella, a species known to have a relatively high prevalence for detection when the virus is present in a population (Ha et al. 2007).
Other opportunities for non-invasive surveillance include sampling of nest material 2060 for persistent and environmentally stable viruses, such as BFDV (Sikorski et al. 2013).
Involving experts in wildlife disease and diagnostics will aid in designing surveillance that meets the particular logistical, statistical and financial needs of the program and species at hand.
Comprehensive health and disease screening programs are essential for captive 2065 breeding or rehabilitation facilities, to ensure they remain free of diseases that may be spread to the wild (Jackson et al. 2014a). To achieve this, expert opinion should be sought from wildlife veterinarians as to prioritisation of screening and appropriate use of available diagnostic tests. Costs of disease screening should be built into captive programs and translocations/reintroductions, as it is a part of 2070 mitigating the recognised risks of pathogen spread through animal movements
(Sainsbury and Vaughan-Higgins 2012). Further, it is a means by which to safeguard
181 the source and recipient populations, with documented failures to detect known pathogens resulting in spread of disease to naïve wildlife populations with associated mortalities (Walker et al. 2008). Whilst it is important to incorporate disease 2075 screening into wider management and program objectives, it is now considered unacceptable to release an individual that may harm the wider ecosystem into which it is released through disease spread given the currently available knowledge and prevention tools.
There is now undeniable evidence that diseases of wildlife need to be factored into 2080 contemporary conservation planning (Spalding and Forrester 1993, Daszak and
Cunningham 1999, Grogan et al. 2014). Anthropogenic influences on ecosystems and climate are driving disease emergence and spread (Daszak et al. 2001, Gottdenker et al. 2014). The evolving human-wildlife-domestic animal interface is altering disease dynamics and providing an ongoing threat of pathogen spill over and spill back 2085
(Packer et al. 1999, Daszak et al. 2000, Miller et al. 2002) and pathogen evolution
(McCarthy et al. 2007). Finally, the expansion of exotic and invasive species provides both a conduit and reservoir for pathogens, that may enable persistence contrary to basic disease theory on endemic and epidemic fade out of highly virulent pathogens in small populations (de Castro and Bolker 2005, Jackson et al. 2014b). Disease in a 2090 large population with adequate habitat and resources may be tolerated and even
‘normal’. However wildlife species no longer live in ‘normal’ environments, neither are many species fortunate enough to be comprised of resilient, genetically diverse, large populations. Consequently many species, particularly those in small or isolated populations, are at greater risk of extinction by disease (de Castro and Bolker 2005, 2095
182 Heard et al. 2013). Therefore there is a clear need to incorporate health and disease objectives into species recovery and conservation plans to ensure the full suite of current and potential future threats are recognised and addressed.
183 References
Aiello C, Nussear K, Walde A, Esque T, Emblidge P, Sah P, Bansal S, Hudson P. 2014.
Disease dynamics during wildlife translocations: disruptions to the host population and potential consequences for transmission in desert tortoise contact networks.
Animal Conservation 17:27-39.
Alley M, Connolly J, Fenwick S, Mackereth G, Leyland M, Rogers L, Haycock M, Nicol
C, Reed C. 2002. An epidemic of salmonellosis caused by Salmonella Typhimurium
DT160 in wild birds and humans in New Zealand. New Zealand Veterinary Journal
50(5):170-176.
Altizer S, Harvell D, Friedle E. 2003. Rapid evolutionary dynamics and disease threats to biodiversity. Trends in Ecology and Evolution 18:589-596.
Armstrong D, Castro I, Griffiths R. 2007. Using adaptive management to determine requirements of re-introduced populations: the case of the New Zealand hihi.
Journal of Applied Ecology 44:953-962.
Armstrong D, Ewen J. 2013. Consistency, continuity and creativity: long-term studies of population dynamics on Tiritiri Matangi Island. New Zealand Journal of Zoology
37(3):288-297.
Auger A, Nantel J, Meunier N, Harrison J. 1979. Skin acariasis caused by Dermanyssus gallinae (de Geer) an in-hospital outbreak. Canadian Medical Association Journal
17:700-703.
184 Bailey T, Wernery U, Naldo J, Samour J. 1997. Plasma concentration of creatine kinase and lactate dehydrogenase in Houbara Bustards (Chlamydotis undulata macqueenii) immediately following capture. Comparative Haematology International
1:113-116.
Baling M, Ji W, Gartrell B, Brunton DH. 2013. Detection of Salmonella during the translocation of two endemic New Zealand lizard species within the Hauraki Gulf.
New Zealand Journal of Zoology 40(3):249-254.
Baron H, Howe L, Varsani A. 2014. Disease screening of three breeding populations of adult exhibition budgerigars (Melopsittacus undulatus) in New Zealand reveals a high prevalence of a novel polyomavirus and avian malaria infection. Avian Diseases
28(2).
Bellingham P, Towns D, Cameron E, Davis J, Wardle D, Wilmshurst J, Mulder C. 2009.
New Zealand island restoration: seabirds, predators, and the importance of history.
New Zealand Journal of Ecology 34(1):115-136.
Berentsen P, Dijkhuizen A, Oskam A. 1992. A dynamic model for cost-benefit analyses of foot-and-mouth disease control strategies. Preventive Veterinary
Medicine 12(3):229-243.
Berggren Å. 2005. Effect of the blood-sucking mite Ornithonyssus bursa on chick growth and fledging age in the North Island robin. New Zealand Journal of Zoology
29(2):243-250.
BirdLife International. 2013. IUCN 2013 Red List of Threatened Species. Version
2013.2, http://www.iucnredlist.org. Accessed June 25th 2013.
185 Bishop D, Heath A. 1998a. Checklist of ectoparasites of birds in New Zealand.
Surveillance 25:13-31.
Bishop D, Heath A. 1998b. Special Issue: Parasites of birds in New Zealand - checklist of ectoparasites. Surveillance 25(5):13-14.
Blackmore D. 1963. Some observations on Cnemidocoptes pilae, together with its effect on the budgerigar (Melopsittacus undulatus). The Veterinary Record
75(23):592-594.
Boon W-M, Kearvell J, Daugherty M, Chambers GK. 2001. Molecular systematics and conservation of kakariki (Cyanoramphus sp). In D. O. Conservation, (ed.). Science for
Conservation, Wellington, 1173-2946.
Borthwick D. 2005. Threat abatement plan for Psittacine Beak and Feather Disease affecting endangered psittacine species. Department of the Environment and
Heritage, Commonwealth of Australia.
Brennan M, Alley M. 2008. Mortalities in red-crowned parakeets (kakariki),
Cyanoramphus novaezelandiae, translocated to Motuihe Island. Kokako 15(2):49.
Briscoe J, Rosenthal K, Shofer F. 2010. Selected complete blood cell count and plasma protein electrophoresis parameters in pet psittacine birds evaluated for illness. Journal of Avian Medicine and Surgery 24(2):131-137.
Burke D, Erica N. 1998. Influence of food abundance, nest-site habitat, and forest fragmentation on breeding ovenbirds. The Auk 115(1):96-104.
186 Businga N, Langenber J, Carlson L. 2007. Successful treatment of capture myopathy in three wild greater sandhill cranes (Grus canadensis tabida). Journal of Avian
Medicine and Surgery 21(4):294-298.
Cade T. 2007. Exposure of California Condors to lead from spent ammunition. The
Journal of Wildlife Management 71(7):2125-2133.
Campbell D. 2011. Seedling recovery on Hauturu/Little Barrier Island after eradication of Pacific Rats Rattus exulans. Department of Conservation Wellington, pp.
Campbell T, Ellis C. 2007. Avian and Exotic Animal Hematology and Cytology. 3rd Ed.
Blackwell Publishing, Iowa.
Capitelli R, Crosta L. 2013. Overview of Psittacine Blood Analysis and Comparative
Retrospective Study of Clinical Diagnosis, Hematology and Blood Chemistry in
Selected Psittacine Species. Veterinary Clinics of North America: Exotic Animal
Practice 16:71-120.
Carlson BL, Roher DP, Nielsen SW. 1982. Notoedric mange in gray squirrels (Sciurus carolinensis). Journal of Wildlife Diseases 18(3):347-348.
Carlson HC, Allen JRD-. 1969. The acute inflammatory reaction in chicken skin: blood cellular response. Avian Diseases 13(4):817-833.
Cencek T. 2003. Prevalence of Dermanyssus gallinae in poultry farms in the Silesia region in Poland. Bulletin Veterinary Institute in Pulawy 47:465-469.
187 Chauve C. 1998. The poultry red mite Dermanyssus gallinae (De Geer, 1778): current situation and future prospects for control. Veterinary Parasitology 79:239-245.
Chirico J, Tauson R. 2002. Traps containing acaricides for the control of Dermanyssus gallinae. Veterinary Parasitology 110:109-116.
Circella E, Pugliese N, Todisco G, Cafiero M, Sparagano O, Camarda A. 2011.
Chlamydia psittaci infection in canaries heavily infested by Dermanyssus gallinae.
Experimental & Applied Acarology 55:329-338.
Claver J, Quaglia A. 2009. Comparative Morphology, Development, and Function of
Blood Cells in Nonmammalian Vertebrates. Journal of Exotic Pet Medicine 18(2):87-
97.
Collings D, Collings B, Julian L, Kurenbach B, Varsani A. 2015. Genome sequences of
Beak and Feather Disease Virus in urban Rainbow Lorikeets (Trichoglossus haematodus). Genome Announcements 3(2).
Cornish T, Linders M, Little S, Vander Haegen W. 2001. Notoedric mange in Western
Gray Squirrels from Washington. Journal of Wildlife Diseases 37(3):630-633.
Cunningham A. 1996. Disease risks of wildlife translocations. Conservation Biology
10(2):349-353.
Curtis C. 2012. Canine sarcoptic mange (sarcoptic acariasis, canine scabies).
Companion Animal 17:32-26.
Dabert J, Dabert M, Gal AF, Miclâus V, Mihalca AD, Sándor AD. 2013.
Multidisciplinary analysis of Knemidocoptes jamaicensis parasitising the Common
188 Chaffinch, Fringilla coelebs: proofs for a multispecies complex? Parasitology
Research 112(6):2373-2380
Dabert J, Mihalca AD, Sándor AD. 2011. The first report of Knemidocoptes intermedius Fain et Macfarlane, 1967 (Acari: Astigmata) in naturally infected
European birds. Parasitology Research 109:237-240.
Dacie J, Lewis S. 1995. Practical Haematology. 8th Ed. Churchill-Livingstone, UK.
Daszak P, Cunningham A. 1999. Extinction by infection. Trends in Ecology and
Evolution 14(3):83-84.
Daszak P, Cunningham A, Hyatt A. 2000. Emerging infectious diseases of wildlife - threats to biodiversity and human health. Science 287:443-449.
Daszak P, Cunningham A, Hyatt A. 2001. Anthropogenic environmental change and the emerging infectious diseases in wildlife. Acta Tropica 78:103-116.
Dawe M. 1979. Behaviour and ecology of the red-crowned kakariki (Cyanoramphus novaezelandiae) in relation to management, Department of Zoology, Auckland
University, Auckland. de Castro F, Bolker B. 2005. Mechanisms of disease-induced extinction. Ecology
Letters 8:117-126.
Di Palma A, Giangaspero A, Cafiero M, Germinara G. 2012. A gallery of the key characters to ease identification of Dermanyssus gallinae (Acari: Gamasida:
Dermanyssidae) and allow differentiation from Ornithonyssus sylviarum (Acari:
Gamasida: Macronyssidae). Parasites and Vectors 5(104).
189 Dickens M, Delehanty D, Romero L. 2010. Stress: an inevitable component of animal translocation. Biological Conservation 143:1329-1341.
Dobson A, Lyles A. 2000. Black-footed ferret recovery. Science 288(5468):985-988.
Doneley RJT. 2003. Acute Beak and Feather Disease in juvenile African Grey parrots - an uncommon presentation of a common disease. Australian Veterinary Journal
81(4):206-207.
Dudley J. 2008. Public health and epidemiological considerations for avian influenza risk mapping and risk assessment. Ecology and Society 13(2):21.
Eastwood J, Berg M, Ribot R, Raidal S, Buchanan K, Walder K. 2014. Phylogenetic analysis of beak and feather disease virus across a host ring-species complex. PNAS
11(39):14153-14158.
Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic acids research 32(5):1792-1797.
Eidson M, Komar N, Sorhage F, Nelson R, Talbot T, Motastashari F, McLean R. 2001a.
Crow deaths as sentinel surveillance system for West Nile virus in the Northeastern
United States, 1999. Emerging Infectious Diseases 7(4):615-620.
Eidson M, Miller J, Kramer L, Cherry B, Hagiwara Y. 2001b. Dead crow densities and human cases of West Nile Virus, New York State, 2000. Emerging Infectious Diseases
7(4):662-664.
190 Elliot G, Dilks P, O'Donnell C. 1996. The ecology of yellow-crowned parakeets
(Cyanoramphus auriceps) in Nothofagus forest in Fiordland, New Zealand. New
Zealand Journal of Zoology 23(3):249-265.
Ewen J, Thorogood R, Brekke P, Cassey P, Karadas F, Armstrong D. 2009. Maternally invested carotenoids compensates costly ectoparasitism in the hihi. PNAS
106(31):12798-12802.
Ewen J, Thorogood R, Nicol C, Armstrong D, Alley M. 2007. Salmonella Typhimurium in Hihi, New Zealand. Emerging Infectious Diseases 13(5):788-789.
Fain A. 1966. Procnemidocoptes janssensi g.n., sp. n. et Neocnemidocoptes g.n.
Remaniement des Knemidokoptidae, parasites cuticoles des Oiseaux (Acarina:
Sarcoptiformes). Revue de Zoologie et de Botanique Africaines 73:390-396.
Fain A. 1974. Notes sur les Knemidokoptidae avec description de taxa nouveaux.
Acarologia 16:182-188.
Fain A, Elsen P. 1967. Les acariens de la famille Knemidocoptidae producteurs de gale ches les oiseaux. Acta Zool Pathol Antverp 47:3-145.
Fain A, Lukoschus F. 1979. Two new parasitic mites (Acari, Astigmata) from the skin of Australian vertebrates. Mitteilungen aus dem Hamburgischen Zoologischen
Museum Institut 76:387-393.
Fletcher OJ. 2008. Avian Histopathology. 3rd edition Ed. American Association of
Avian Pathologists, Raleigh, North Carolina.
Forshaw JM. 2010. Parrots of the World. CSIRO, Collingwood.
191 French N, Midwinter A, Holland B, Collins-Emerson J, Pattison R, Colles F, Carter P.
2009. Molecular epidemiology of Campylobacter jejuni isolates from wild-bird fecal material in children's playgrounds. Applied and Environmental Microbiology
75(3):779-783.
Fridolfsson A, Ellegren H. 1999. A simple and universal method for molecular sexing of non-ratite birds. Journal of Avian Biology 20:116-121.
Galbraith M, Cooper H. 2013. Tiritiri Matangi - an overview of 25 years of ecological restoration. New Zealand Journal of Zoology 37(3):258-260.
Garrett L, Haramoto F. 1967. A catalog of Hawaiian Acarina. Proceedings Hawaiian
Entomological Society 19:381-414.
Gartrell B, French N, Howe L, Nelson N, Houston M, Burrows E, Russell J, Anderson S.
2013. First detection of Chlamydia psittaci from a wild native passerine bird in New
Zealand. New Zealand Veterinary Journal 61(3):174-176.
Gartrell B, Stone M, King C, Wang J. 2003. An outbreak of disease caused by psittacinepoxvirus in rosellas. Surveillance 30(3):11-13.
Gaudioso JM, LaPointe DA, Hart PJ. 2009. Knemidokoptic mange in Hawaiian amakihi
(Hemignathus virens) on the island of Hawaii. Journal of Wildlife Diseases 45(2):497-
501.
Gilardi KV, Gilardi JD, Frank A, Goff ML. 2001. Epidermoptid mange in Laysan albatross fledglings in Hawaii. Journal of Wildlife Diseases 37(1):185-188.
192 Godfrey S, Moore J, Nelson N, Bull C. 2010. Social network structure and parasite infection patterns in a territorial reptile, the tuatara (Sphenodon punctatus).
International Journal for Parasitology 40:1575-1585.
Gottdenker N, Streicker D, Faust C, Carroll C. 2014. Anthropogenic land use change and infectious diseases: a review of the evidence. EcoHealth.
Graham M, Veitch D, Aguilar G, Galbraith M. 2013. Monitoring terrestrial bird populations on Tiritiri Matangi Island, Hauraki Gulf, New Zealand, 1987–2010. New
Zealand Journal of Zoology 37(3):359-369.
Grear D, Luong L, Hudson P. 2013. Network transmission inference: Host behaviour and parasite life cycle make social networks meaningful in disease ecology.
Ecological Applications 23(8):1906-1914.
Greene T. 1998. Foraging ecology of the red-crowned parakeet (Cyanoramphus novaezelandiae novaezelandiae) and yellow-crowned parakeet (Cyanoramphus auriceps auriceps) on Little Barrier Island, Hauraki Gulf, New Zealand. New Zealand
Journal of Ecology 22(2):161-171.
Greene T. 2003. Breeding biology of red-crowned parakeets (Cyanoramphus novaezelandiae novaezelandiae) on Little Barrier Island, Hauraki Gulf, New Zealand.
Notornis 50:83-99.
Grogan L, Berger L, Rose K, Grillo V, Cashins S, Skerratt L. 2014. Surveillance for
Emerging Biodiversity Diseases of Wildlife. PLOS Pathogens 10(5).
193 Guindon Sp, Dufayard J-Fo, Lefort V, Anisimova M. 2010. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of
PhyML 3.0. Systematic Biology 59(3):307-321.
Ha H, Alley M, Cahill J, Howe L. 2009. The prevalence of psittacine beak and feather disease virus infection in native parrots in New Zealand. New Zealand Veterinary
Journal 57(1):50-52.
Ha H, Anderson I, Alley, Springett B. 2007. The prevalence of beak and feather disease virus infection in wild populations of parrots and cockatoos in New Zealand.
New Zealand Veterinary Journal 55(5):235-238.
Harkins GW, Martin DP, Christoffels A, Varsani A. 2014. Towards inferring the global movement of beak and feather disease virus. Virology 450-451:24-33.
Harrington D, Robinson K, Guy J, Sparagano O. 2010. Characterization of the immunological response to Dermanyssus gallinae infestation in domestic fowl.
Transboundary and Emerging Diseases 57(1-2):107-110.
Harrison G, Lightfoot T. 2006. Clinical Avian Medicine. Spix Publishing, Palm Beach
Florida.
Heard M, Smith K, Ripp K, Berger M, Chen J, Dittmeier J, Goter M, McGarvey S, Ryan
E. 2013. The threat of disease increases as species move toward extinction.
Conservation Biology 27(6):1378-1388.
Heath A. 2010. Checklist of ectoparasites of birds in New Zealand: additions and corrections. Surveillance 37:12-17.
194 Heber S, Briskie JV. 2010. Population bottlenecks and increased hatching failure in endangered birds. Conservation Biology 24(6):1674-1678.
Hess M, Scope A, Heincz U. 2004. Comparitive sensitivity of polymerase chain reaction diagnosis of psittacine beak and feather disease on feather samples, cloacal swabs and blood from budgerigars (Melopsittacus undulatus, Shaw 1805). Avian
Pathology 33(5):477-481.
Holden J. 1997. Trade in CITES-listed birds to and from New Zealand. TRAFFIC
Oceania, Sydney.
Holz P, Beveridge I, Ross T. 2005. Knemidocoptes intermedius in wild superb lyrebirds
(Menura novaehallandiae). Australian Veterinary Journal 83(6):374-375.
Howe L, Castro I, Schoener E, Hunter S, Barraclough R, Alley M. 2012. Malaria parasites (Plasmodium spp.) infecting introduced, native and endemic New Zealand birds. Parasitology Research 110:913-923.
Hrubec T, Whichard J, Larsen C, William Pierson F. 2002. Plasma versus Serum:
Specific differences in biochemical analyte values. Journal of Avian Medicine and
Surgery 16(2):101-105.
Ihaka R, Gentleman R. 1996. R: A Language for Data Analysis and Graphics. Journal of
Computational and Graphical Statistics 5(3):299-314.
Innes J, Kelly D, McOverton J, Gillies C. 2010. Predation and other factors currently limiting New Zealand forest birds. New Zealand Journal of Ecology 34(1).
195 ISIS. 2013. ISIS Physiological Reference Intervals for Captive Wildlife: A CD-ROM.
International Species Information System, USA.
Jackson B, Harvey C, Galbraith J, Robertson M, Warren K, Holyoake C, Julian L,
Varsani A. 2014a. Clinical beak and feather disease virus infection in wild juvenile eastern rosellas of New Zealand; biosecurity implications for wildlife care facilities.
New Zealand Veterinary Journal.
Jackson B, Lorenzo A, Theuerkauf J, Barnaud A, Duval T, Guichard P, Bloc H, Baouma
A, Stainton D, Kraberger S, Murphy S, Clark N, Dillon C, Knight T, Varsani A. 2014b.
Preliminary surveillance for beak and feather disease virus in wild parrots of New
Caledonia: implications of a reservoir species for Ouvea Parakeets. Emu 114:283-
289.
Jaensch SM, Raidal SR, Hobbs R. 2003. Knemidocoptes intermedius in a wild currawong (Strepera graculina). Australian Veterinary Journal 81(7):411.
Jakob-Hoff R. 2002. Avian Health Surveillance Project. Auckland Zoo, Auckland.
Jakob-Hoff R. 2010. Diagnosticians upskill in avian haemoparasite recognition.
VetScript.
Jakob-Hoff R, MacDiarmid S, Lees C, Miller P, Travis D, Kock R. 2014. Manual of
Procedures for Wildlife Disease Risk Analysis. World Organisation for Animal Health,
Paris.
196 Jansen J. 1957. Knemidokoptes pilae Lavoipierre and Griffiths, 1951, found on
Palaeornis torquatus and Ara sp; with a special note on the bases of the adscutal setae. Parasitology 47:160-164.
Jensen T, Pernasetti F, Durrant B. 2003. Conditions for rapid sex determination in 47 avian species by PCR of genomic DNA from blood, shell-membrane blood vessels, and feathers. Zoo Biology 22:561-571.
Jessup D, Chechowitz M, Ames J, Harris M, Worcester K, Paradies D, Mazet J. 2001.
The decline of the southern sea otter (Enhydra lutris nereis): Is it an indicator of contaminant and pathogen pollution, anthropogenic interference, and declining marine ecosystem health? In: Proceedings of the Proceedings AAZV and IAAAM Joint
Conference pp. 13-17.
Johnson N, Mueller J. 2002. Updating the accounts: Global mortality of the 1918-
1920 "Spanish" Influenza Pandemic. Bulletin of the History of Medicine 76(1).
Jones G, Jacobs D, Kunz T, Willig M, Racey P. 2009. Carpe noctem: the importance of bats as bioindicators. Endangered Species Research 8:93-115.
Jones K, Patel N, Levy M, Storeygard A, Balk D, Gittleman J, Daszak P. 2008. Global trends in emerging infectious diseases. Nature 451:990-994.
Julian L, Lorenzo A, Chenuet J-P, Bonzon M. 2012. Evidence of multiple introductions of beak and feather disease virus into the Pacific islands of Nouvelle-Caledonie (New
Caledonia). Journal of General Virology 93:2466-2472.
197 Julian L, Piasecki T, Chrzastek K, Walters M. 2013. Extensive recombination detected among beak and feather disease virus isolates from breeding facilities in Poland.
Journal of General Virology 94:1086-1095.
Karesh W. 1993. Cost evaluation of infectious disease monitoring and screening programs for wildlife translocation and reintroduction. Journal of Zoo and Wildlife
Medicine 24(3):291-295.
Karesh W, Cook R, Bennett E, Newcomb J. 2005. Wildlife trade and global disease emergence. Emerging Infectious Diseases 11(7):1000-1002.
Khalesi B, Bonne N, Stewart M, Sharp M, Raidal S. 2005 A comparison of haemagglutination, haemagglutination inhibition and PCR for the detection of psittacine beak and feather disease virus infection and a comparison of isolates obtained from loriids. Journal of General Virology 86 (11 ):3039-3046
Kirkwood J. 1998. Population density and infectious disease at bird tables. The
Veterinary Record 142:468.
Kirmse P. 1966. Cnemidocoptic mite infestations in wild birds. Bulletin Wildlife
Disease Association 2:86-99.
Kitulagodage M, Buttemer W, Astheimer L. 2011. Adverse effects of fipronil on avian reproduction and development: maternal transfer of fipronil to eggs in zebra finch
Taeniopygia guttata and in ovo exposure in chickens Gallus domesticus.
Ecotoxicology 20:653-660.
198 Kock R, Woodford M, Rossiter P. 2010. Disease risks associated with the translocation of wildlife. Rev sci tech Off int Epiz 29(2).
Kolodziej-Sobocińska M, Zalewski A, Kowalczyk R. 2014. Sarcoptic mange vulnerability in carnivores of the Białowieża Primeval Forest, Poland: underlying determinant factors. Ecological Research 29:237-244.
Koski MA. 2002. Dermatologic diseases in psittacine birds: An investigational approach. Seminars in Avian and Exotic Pet Medicine Anatomy and Disorders of the
Integument 11(3):105-124.
Krantz G, Walter D. 2009. A manual of Acarology. 3rd Ed. Texas Tech University
Press, USA.
Krebs E. 1998. Breeding biology of crimson rosellas (Platycercus elegans) on Black
Mountain, Australian Capital Territory. Australian Journal of Zoology 46:119-136.
Kundu S, Faulkes CG, Greenwood AG, Jones CG. 2012. Tracking viral evolution during a disease outbreak: the rapid and complete selective sweep of a circovirus in the endangered Echo parakeet. Journal of Virology 86(9):5221-5229.
Kutzer E. 1964. Knemidocoptes-Raude bei Ziervogeln. Wien tierarztl Mschr 51:36-43.
Ladds. 2009. The Pathology of Australian native wildlife. CSIRO Publishing,
Collingwood.
Lafferty K, Gerber L. 2002. Good medicine for conservation biology: the intersection of epidemiology and conservation theory. Conservation Biology 16(3):593-604.
199 Lafferty K, Smith K, Torchin M, Dobson A, Kuris A. 2005. The Role of Infectious
Diseases in Natural Communities.In: Species invasions: insights into ecology, evolution and biogeography, D. Sax, J. Stachowicz, AND S. Gaines, (editors.). Sinauer
Associates Incorporated, Sunderland, USA.
Lavoipierre M, Griffiths R. 1951. A preliminary note on a new species of
Cnemidocoptes (Acarina) causing scaly leg in a budgerigar (Melopsittacus undulatus) in Great Britain Annals of Tropical Medical Parasitology 45:253-254.
Lawson B, Hughes L, Pennycott T, Simpson V, John S, Peck K, Toms M, Bennett M,
Kirkwood J, Cunningham A, Robinson R, Neimanis A, Handeland K, Isomursu M,
Agren E, Hamnes I, Tyler K, Chantrey J. 2011. Evidence of spread of the emerging infectious disease, finch trichomonosis, by migrating birds. EcoHealth 8(2):143-153.
Le Souef A. 2012. Development and improvement of clinical tools for rehabilitating endangered Black Cockatoos (Calyptorhynchus spp.) back to the wild, Murdoch
University, Murdoch, WA, Australia.
Le Souef A, Holyoake C, Vitali S, Warren K. 2013. Hematologic and plasma biochemical reference values for three species of black cockatoos (Calyptorhynchus species). Journal of Avian Medicine and Surgery 27(1):14-28.
Legge S, Murphy S, Igag P, Mack A. 2004. Territoriality and density of an Australian migrant, the Buff-breasted Paradise Kingfisher, in the New Guinean non-breeding ground. Emu 104:15-20.
Lehmann T. 1993. Ectoparasites: Direct impact on host fitness. Parasitology Today
9(1):8-12.
200 Lei B, Green J, Pichegru L. 2014. Extreme microclimate conditions in artifical nests for endangered African Penguins. Bird Conservation International 24(2):201-213.
León V, Ruiz de Ibáñez M, Cubero-Pablo M, Ortiz J, Espinosa J, Pérez L, Simón M,
Alonso F. 1999. Sarcoptic mange in Iberian ibex from Spain. Journal of Wildlife
Diseases 35:647-659.
Literák I, Śmíd B, Valíček L. 2003. Papillomatosis in chaffinches (Fringella coelebs) in the Czech Republich and Germany. Vet Med Czech 48:169-173.
Lloyd-Smith JO, Cross PC, Briggs CJ, Daugherty M, Getz WM, Latto J, Sanchez MS,
Smith AB, Swei A. 2005. Should we expect population thresholds for wildlife disease?
TRENDS in Ecology and Evolution 20(9):511-519.
Losson B. 2012. Sheep psoroptic mange: An update. Veterinary Parasitology 189:39-
43.
Low M, Alley MR, Scott I. 2007. Pruritic Facial Dermatitis in a Population of Free-
Living Stitchbirds. Journal of Wildlife Diseases 43(2):262-268.
Low M, Eason D, Elliot G, McInnes K, Paul-Murphy J. 2006. Hematologic and biochemical reference ranges for the kakapo (Strigops habroptilus): generation and interpretation in a field-based wildlife recovery program. Journal of Avian Medicine and Surgery 20(2):80-88.
Maddison WP, Maddison DR. 2011. Mesquite: a modular system for evolutionary analysis. pp.
201 Mainka SA, Melville DS, Galsworthy A, Black SR. 1994. Knemidocoptes sp. on wild passerines at the Mai Po Nature Reserve, Hong Kong. Journal of wildlife diseases
30(2):254-256.
Makan T. 2009. A critique of disease-screening requirements for community- managed species translocations: A case study of hihi (Notiomystis cincta). Royal
Forest and Bird Society of New Zealand, Auckland.
Marangi M, Morelli V, Pati S, Carmada A, Assunta Cafiero M, Giangaspero A. 2012.
Acaricide residues in laying hens naturally infested by red mite Dermanyssus gallinae. PLoS ONE 7(2):e31795.
Mason R, Fain A. 1988. Knemidocoptes intermedius identified in forest ravens
(Corvus tasmanicus). Australian Veterinary Journal 65(8):260.
Massaro M, Ortiz-Catedral L, Julian L, Galbraith J, Kurenbach B, Kearvell J, Kemp J,
Hal J, Elkington S, Taylor G, Greene T, Wetering J, Wetering M, Pryde M, Dilks P,
Heber S, Steeves T, Walters M, Shaw S, Potter J, Farrant M, Brunton D, Hauber M,
Jackson B, Bell P, Moorhouse R, McInnes K, Varsani A. 2012. Molecular characterisation of beak and feather disease virus (BFDV) in New Zealand and its implications for managing an infectious disease. Archives of Virology 157:1651-1663.
Mathews F, Moro D, Strachan R, Gelling M, Buller N. 2006. Health surveillance in wildlife reintroductions. Biological Conservation 131(2):338-347.
McCarthy A, Shaw M, Goodman S. 2007. Pathogen evolution and disease emergence in carnivores. Proceedings of the Royal Society Biological Sciences 274:3165-4174.
202 McDonald D, Jaensch SM, Harrison G, Gelis S, Brennan D, Sacks P, Bernardo D. 2010.
Health and nutritional status of wild Australian Psittacine birds: an evaluation of plasma and hepatic mineral levels, plasma biochemical values, and fecal microflora.
Journal of Avian Medicine and Surgery 24(4):288-298.
McHalick O. 1998. Translocation database summary. Biodiversity Recovery Unit,
DOC, Wellington.
ME. 1997. State of the New Zealand Environment. Ministry for the Environment,
Wellington.
ME. 2000. New Zealand Biodiversity Strategy. Ministry for the Environment,
Wellington.
ME. 2007. Environment New Zealand 2007. Ministry for the Environment,
Wellington.
Meyer-K hling B, Heine J, M ller-Lindloff J, Pfister K. 2007. Epidemiology of
Dermanyssus gallinae and acaricidal efficacy of Phoxim 50% in alternative housing systems during the laying period of hens. Parasitology Research 101:1-12.
Miller D, Taton-Allen G, Campbell T. 2004. Knemidokoptes in a Swainson's Hawk,
Buteo Swainsoni. Journal of Zoo and Wildlife Medicine 35(3):400-402.
Miller M, Gardner I, Kreuder C, Paradies D, Worcester K, Jessup D, Dodd E, Harris M,
Ames J, Packham A, Conrad P. 2002. Coastal freshwater runoff is a risk factor for
Toxoplasma gondii infection of southern sea otters (Enhydra lutris nereis).
International Journal for Parasitology 32:997-1006.
203 Mironov SV, Bochkov AV, Fain A. 2005. Phylogeny and evolution of parasitism in feather mites of the families Epidermoptidae and Dermationidae (Acari: Analgoidea).
Zoologischer Anzeiger - A Journal of Comparative Zoology 243(3):155-179.
Mitchell E, Johns J. 2008. Avian Haematology and Related Disorders. Veterinary
Clinics of North America: Exotic Animal Practice 11:501-522.
Møller A. 1990. Effects of parasitism by a haematophagous mite on reproduction in the barn swallow. Ecology 71(6):2345-2357.
Møller A. 2002. Temporal change in mite abundance and its effect on barn swallow reproduction and sexual selection. Journal of Evolutionary Biology 15:495-504.
Morishita T, Johnson G, Johnson G, Thilsted J, Promsopone B, Newcomer C. 2005.
Scaly-leg mite infestation associated with digit necrosis in bantam chickens (Gallus domesticus). Journal of Avian Medicine and Surgery 19(3):230-233.
Mount Bruce Native Bird Reserve. 1974. Disposal of Surplus Stock. In W. B.
Controller, (ed.), Masterton.
Muhire B, Martin DP, Brown JK, Navas-Castillo J, Moriones E, Zerbini FM, Rivera-
Bustamante R, Malathi VG, Briddon RW, Varsani A. 2013. A genome-wide pairwise- identity-based proposal for the classification of viruses in the genus Mastrevirus
(Geminiviridae). Archives of Virology 158(6):1411-1424.
Müllen G, Durden L. 2009. Medical and Veterinary Entomology. Elsevier, Amsterdam.
204 Mwacalimba K, Mumba C, Munyeme M. 2013. Cost benefit analysis of tuberculosis control in wildlife-livestock interface areas of Southern Zambia. Preventive
Veterinary Medicine 110:274-279.
Myers N, Mittermeier RA, Mittermeier CG, Fonseca GABd, Kent J. 2000. Biodiversity hotspots for conservation priorities. Nature 403(6772):853-858.
Nellemann C, Corcoran E. 2010. Dead Planet, Living Planet - Biodiversity and
Ecosystem Restoration for Sustainable Development. A Rapid Response Assessment.
UNEP. GRID-Arendal, Norway.
Newton LG, O'Sullivan PJ. 1956. A note on the occurrence of the mite Cnemidocoptes pilae on budgerigars in Queensland. Australian Veterinary Journal 32(4):89-90.
NHMRC. 2013. Australian code for the care and use of animals for scientific purposes.
NHMRC (ed.), Canberra.
NIWA. 2012. Seasonal Climate Summary: Summer 2011/12.
NIWA. 2013. Seasonal Climate Summary: Summer 2012/13.
Nusser SM, Clark WR, Otis DL, Huang L. 2008. Sampling considerations for disease surveillance in wildlife populations. Journal of Wildlife Management 72(1):52-60.
O'Brien C, van Riper III C, Myers D. 2009. Making reliable decisions in the study of wildlife diseases: using hypothesis tests, statistical power, and observed effects.
Journal of Wildlife Diseases 45(3):700-712.
O'Brien DJ. 1999. Treatment of psoroptic mange with reference to epidemiology and history. Veterinary Parasitology 83:177-185.
205 O'Grady B. 1960. Knemidokoptic mange of the budgerigar (Melopsittacus undulatus) in New Zealand. New Zealand Veterinary Journal 8(6):118.
Oldham J, Beresford-Jones W. 1954. Observations on the occurrence of
Cnemidocoptes pilae in budgerigars and a parakeet. British Veterinary Journal
110(29):29-30.
Oleaga A, Casais R, Prieto J, Gortázar C, Balseiro A. 2012. Comparative pathological and immunohistochemical features of sarctoptic mange in five sympatric wildlife species in Northern Spain. European Journal Wildlife Research 58:997-1000.
Ortiz-Catedral L. 2010. Homing of a red-crowned parakeet (Cyanoramphus novaezelandiae) from Motuihe Island to Little Barrier Island, New Zealand. Notornis
57:48-49.
Ortiz-Catedral L, Brunton DH. 2008. Clutch parameters and reproductive success of a translocated population of red-crowned parakeet (Cyanoramphus novaezelandiae).
Australian Journal of Zoology 56:389-393.
Ortiz-Catedral L, Brunton DH. 2010a. Nesting sites and nesting success of reintroduced red-crowned parakeets (Cyanoramphus novaezelandiae) on Tiritiri
Matangi Island, New Zealand. New Zealand Journal of Zoology 36:1-10.
Ortiz-Catedral L, Brunton DH. 2010b. Success of translocations of red-fronted parakeets Cyanoramphus novaezelandiae novaezelandiae from Little Barrier Island
(Hauturu) to Motuihe Island, Auckland, New Zealand. Conservation Evidence 7:21-26.
206 Ortiz-Catedral L, Brunton DH, Kearvell J. 2012. Population increase of critically endangered Malherbe's parakeet Cyanoramphus malherbi introduced to Maud
Island, New Zealand. Conservation Evidence 9:54-57.
Ortiz-Catedral L, Hauber M, Brunton DH. 2013. Growth and survival of nestlings in a population of red-crowned parakeets (Cyanoramphus novaezelandiae) free of introduced mammalian nest predators on Tiritiri Matangi Island, New Zealand. New
Zealand Journal of Zoology 37(3):370-378.
Ortiz-Catedral L, Ismar S, Baird K, Ewen J, Hauber M, Brunton D. 2009a. No evidence of Campylobacter, Salmonella and Yersinia in free-living populations of the red- crowned parakeet (Cyanoramphus novaezelandiae). New Zealand Journal of Ecology
36:379-383.
Ortiz-Catedral L, Kurenbach B, Massaro M, Mcinnes K, Brunton DH, Hauber ME,
Martin DP, Varsani A. 2010. A new isolate of beak and feather disease virus from endemic wild red-fronted parakeets (Cyanoramphus novaezelandiae) in New Zealand
AN - 212378819; 20180139. Archives of Virology 155(4):613-620.
Ortiz-Catedral L, McInnes K, Hauber ME, Brunton DH. 2009b. First report of beak and feather disease virus (BFDV) in wild Red-fronted Parakeets (Cyanoramphus novaezelandiae) in New Zealand. Emu 109(3):244-247.
Ortiz-Catedral L, Prada D, Gleeson D, Brunton D. 2011. Avian malaria in a remnant population of red-fronted parakeets on Little Barrier Island, New Zealand. New
Zealand Journal of Zoology 38(3):261-268.
207 Orton D, Warren L, Wilkinson J. 2000. Avian mite dermatitis. Clinical and
Experimental Dermatology 25:129-131.
Owen J, Delany M, Cardona C, Bickford A, Mullens B. 2009. Host inflammatory response governs fitness in an avian ectoparasite, the northern fowl mite
(Ornithonyssus sylviarum). International Journal for Parasitology 39:789-799.
Packer C, Altizer S, Appel M, Brown E, Martenson J, O'Brien S, Roelke-Parker M,
Hofmann-Lehmann R, Lutz H. 1999. Viruses of the Serengeti: Patterns of infection and mortality in African lions. Journal of Animal Ecology 68:1161-1178.
Pain D, Martins T, Boussekey M, Diaz S, Downs C, Ekstrom J, Garnett S, Gilardi JD,
McNiven D, Primot P, Rouys S, Saoumoé M, Symes C, Tamungang S, Theuerkauf J,
Villafuerte D, Verfailles L, Widmann P, Widmann I. 2006. Impact of protection on nest take and nesting success of parrots in Africa, Asia and Australasia. Animal
Conservation 9:322-330.
Parker K. 2013. Avian translocations to and from Tiritiri Matangi 1974-2013. New
Zealand Journal of Ecology 37(3):282-287.
Parker K, Brunton D, Jakob-Hoff R. 2006. Avian translocations and disease; implications for New Zealand conservation. Pacific Conservation Biology 12(2):155-
162.
Pass DA, Perry RA. 1984. The pathology of psittacine beak and feather disease.
Australian Veterinary Journal 61(3):69-74.
208 Paz Nava M, Veiga J, Puerta M. 2001. White blood cell counts in house sparrows
(Passer domesticus) before and after moult and after testosterone treatment.
Canadian Journal of Zoology 79(1):145-148.
Pence D, Windberg L. 1994. Impact of a sarcoptic mange epizootic on a coyote population. The Journal of Wildlife Management 58:624-633.
Pence DB, Cole RA, Brugger KE, Fischer JR. 1999. Epizootic podoknemidokoptiasis in
American robins. Journal of Wildlife Diseases 35(1):1-7.
Pence DB, Windberg LA, Pence BC, Sprowls RD-. 1983. The epizootiology and pathology of sarcoptic mange in coyotes, Canis latrans, from South Texas. The
Journal of Parasitology 69(6):1100-1115.
Peters A, Patterson EI, Baker BGB, Holdsworth M, Sarker S, Ghorashi SA, Raidal SR.
2014. Evidence of psittacine beak and feather disease virus spillover into wild critically endangered orange-bellied parrots (Neophema chrysogaster) Journal of
Wildlife Diseases 50(2).
Polkanov A. 2006-2009. Rainbow Lorikeet Eradication. Department of Conservation
Auckland.
Polo F, Peinado V, Viscor G, Palomeque J. 1998. Hematologic and plasma chemistry values in captive psittacine birds. Avian diseases 42(3):523-535.
Posada D. 2008. jModelTest: Phylogenetic model averaging. Molecular Biology and
Evolution 25(7):1253-1256.
209 Raidal SR. 1995. Viral skin diseases of birds. Seminars in Avian and Exotic Pet
Medicine Viral Diseases 4(2):72-82.
Raidal SR, Cross GMD-. 1995. Acute necrotizing hepatitis caused by experimental infection with psittacine beak and feather disease virus. Journal of Avian Medicine and Surgery 9(1):36-40.
Raidal SR, McElnea CL, Cross GM. 1993a. Seroprevalence of psittacine beak and feather disease in wild psittacine birds in New South Wales. Australian Veterinary
Journal 70(4):137-139.
Raidal SR, Sabine M, Cross GM. 1993b. Laboratory diagnosis of psittacine beak and feather disease by haemagglutination and haemagglutination inhibition. Australian
Veterinary Journal 70:133-137.
Rajabzadeh M, Razavi SM, Homayounimehr AR. 2008. Depluming itch in native fowls in the Iran, a potential risk for commercial chickens. Tropical Animal Health
Production 40.
Rao S, Ghafoor M, Narsapur V. 1967. Cnemidocoptes pilae (Lavoipierre and Griffiths
1951) a causative agent of scaly leg in budgerigar (Melopsittacus undulatus) in India together with the description of the larval forms. The Indian Veterinary Journal
44(3):206-208.
Regnard G, Rutledge S, Martin R, Hitzeroth I, Rybicki E. 2014a. Beak and feather disease virus: correlation between viral load and clinical signs in wild Cape parrots
(Poicephalus robustus) in South Africa. Archives of Virology.
210 Regnard GL, Rutledge S, Martin R, Hitzeroth I, Rybicki E. 2014b. Beak and feather disease viruses circulating in Cape parrots (Poicepahlus robustus) in South Africa.
Archives of Virology.
Riddoch P, Raidal SR, Cross GM. 1996. Psittacine circovirus antibody detection and an update on the methods for diagnosis of psittacine beak and feather disease.
Australian Veterinary Practitioner 26:134.
Rimmer A. 2009. Tiritiri Matangi: A model of conservation. Random House Books,
Auckland.
Ritchie PA, Anderson IL, Lambert DM. 2003. Evidence for specificity of psittacine beak and feather disease viruses among avian hosts. Virology 306(1):109-115.
Robertson H, Dowding J, Elliot G, Hitchmough R, Miskelly C, O'Donnell C, Powlesland
R, Sagar P, Scofield R, Taylor G. 2012. Conservation status of New Zealand birds,
2012. Department of Conservation, Wellington.
Robino P, Grego E, Rossi G, Bert E, Tramuta C, Stella M, Bertoni P, Nebbia P. 2014.
Molecular analysis and associated pathology of beak and feather disease virus isolated in Italy from young Congo African grey parrots (Psittacus erithacus) with an
"atypical peracute form" of the disease. Avian Pathology 43(4):333-344.
Robinson R, Hutchinson O, John S, Pennycott T, Perkins M, Rowley P, Simpson V,
Tyler K, Cunningham A, Lawson B, Toms M, Peck K, Kirkwood J, Chantrey J,
Clatworthy I, Evans A, Hughes L. 2010. Emerging infectious disease leads to rapid population declines of common British birds. PLoS ONE 5(8).
211 Roe D, Mulliken T, Milledge S, Mosha J, Grief-Gran M. 2002. Making a killing or making a living? Wildlife trade, trade controls and rural livelihoods. IIED, London, pp.
Rosskopf Jr. WJ. 2003. Common conditions and syndromes of canaries, finches, lories and lorikeets, lovebirds, and macaws. Seminars in Avian and Exotic Pet Medicine
Species Syndromes 12(3):131-143.
Roth H, Merz G. 1997. Wildlife Trade Resources: A global account of economic use.
Springer Verlag, Berlin.
Roy L, Chauve C. 2007. Historical review of the genus Dermanyssus Dugès, 1834
(Acari: Mesostigmata: Dermanyssidae). Parasite 14:87-100.
Roy L, Chauve C. 2010. The genus Dermanyssus (Mesostigmata: Dermanyssidae): history and species characterization. In: Proceedings of the Trends in Acarology:
Proceedings of the 12th International Congress, M. Sabelis, AND J. Bruin, editors.
Springer Science.
Ryder J, Miller M, White A, Knell R, Boots M. 2007. Host-parasite population dynamics under combined frequency- and density-dependent transmission Oikos
116:2017-2026.
Ryser-Degiorgis M. 2013. Wildlife health investigations: needs, challenges and recommendations. BMC Veterinary Research 9:223-240.
Sagar P. 1988. Some characteristics of Red-Crowned Parakeets on the Poor Knights
Islands. Notornis 35:1-8.
212 Sainsbury A, Vaughan-Higgins R. 2012. Analyzing disease risks associated with translocations. Conservation Biology 26(3):442-452.
Samour J. 2000. Avian Medicine. 1st Ed. Elsevier Health Sciences, United Kingdom.
Sarker S, Ghorashi S, Forwood J, Bent S, Peters A, Raidal S. 2014a. Phylogeny of beak and feather disease virus in cockatoos demonstrates host generalism and multiple- variant infections within Psittaciformes. Virology 460:72-82.
Sarker S, Ghorashi S, Forwood J, Raidal S. 2014b. Whole-genome sequences of two beak and feather disease viruses in the endangered swift parrot (Lathamus discolor).
Genome Announcements 1(6).
Sarker S, Patterson EI, Peters A, Baker BGB, Forwood JK, Ghorashi SA, Holdsworth M,
Baker R, Murray N, Raidal SR. 2014c. Mutability dynamics of an emergent single stranded DNA virus in a naıve host. PLoS ONE 9(1).
Saunders A, Norton DA. 2001. Ecological restoration at Mainland Islands in New
Zealand. Biological Conservation 99:109-119.
Schloegel L, Daszak P, Cunningham A, Speare R, HIll B. 2010. Two amphibian diseases, chytridiomycosis and ranaviral disease, are now globally notifiable to the
World Organisation for Animal Health (OIE): an assessment. Diseases of Aquatic
Organisms 92:101-108.
Schloegel L, Hero J, Berger L, Speare R, McDonald K, Daszak P. 2006. The decline of the Sharp-snouted Day Frog (Taudactylus acutirostris): The first documented case of extinction by infection in a free-ranging wildlife species? EcoHealth 3:35-40.
213 Schoemaker NJ, Dorrestein GM, Latimer KS, Lumeij JT, Kik MJL, Hage MHvd,
Campagnoli RPD-. 2000. Severe leukopenia and liver necrosis in young African Grey
Parrots (Psittacus erithacus erithacus) infected with psittacine circovirus. Avian
Diseases 44(2):470-478.
Schultz D. 1978. Diseases of parrots. University of Sydney Post-graduate Committee in Veterinary Science Proceedings 36:223-245.
Schulz T, Stewart J, Fowler M. 1989. Knemidokoptes mutans
(Acari:Knemidocoptidae) in a Great-horned Owl (Bubo virginianus). Journal of
Wildlife Diseases 25(3):430-432.
Scott DW, Miller WH. 2011. Equine Dermatology. 2nd edition Ed. Elsevier Saunders,
Missouri.
Seltzer J. 2007. Skin biopsies in mammals. Lab Animal 36(4):23.
Sergeant E. 2015. Epitools epidemiological calculators. AusVet Animal Health
Services and Australian Biosecurity Cooperative Research Centre for Emerging
Infectious Disease. http://epitools.ausvet.com.au.
Shaw D, Grenfell B, Dobson A. 1998. Patterns of macroparasite aggregation in wildlife host populations. Parasitology 117:597-610.
Shoshana R. 1993. Knemidokoptes mites on feathers of the red fronted parakeet.
Proceedings of the European Conference on Avian Medicine and Surgery 329-331.
214 Sikorski A, Kearvell J, Elkington S, Dayaram A, Argaello-Astorga G, Varsani A. 2013.
Novel ssDNA viruses discovered in yellow-crowned parakeet (Cyanoramphus auriceps) nesting material. Archives of Virology 158:1603-1607
Skerratt L, Berger L, Speare R, Cashins S, McDonald K, Phillott A, Hines H, Kenyon N.
2007. Spread of Chytridiomycosis Has Caused the Rapid Global Decline and
Extinction of Frogs. EcoHealth 4:125-134.
Skerratt L, Middleton D, Beveridge I. 1999. Distribution of life cycle stages of
Sarcoptes scabiei var wombati and effects of severe mange on common wombats in
Victoria. Journal of Wildlife Diseases 35(4):633-646.
Sleiman R, Kurban M, Abbas O. 2013. Maximizing diagnostic outcomes of skin biopsy specimens. International Journal of Dermatology 52:72-78.
Spalding M, Forrester D. 1993. Disease monitoring of free-ranging and released wildlife. Journal of Zoo and Wildlife Medicine 24(3):271-280.
Sparagano O, George D, Harrington D, Giangaspero A. 2014. Significance and control of the poultry red mite Dermanyssus gallinae. Annual Review of Entomology 59:447-
466.
Stamp R, Brunton DH, Walter B. 2002. Artificial nest box use by the North Island saddleback: Effects of nest box design and mite infestations on nest site selection and reproductive success. New Zealand Journal of Zoology 29:285-292.
215 Stark K, Regula G, Hernandez J, Knopf L, Fuchs K, Morris R, Davies P. 2006. Concepts for risk-based surveillance in the field of veterinary medicine and veterinary public health: Review of current approaches. BMC Health Services Research 6:20-28.
Storm D, Samuel M, Rolley R, Shelton P, Keuler N, Richards B, Van Deelen T. 2013.
Deer density and disease prevalence influence transmission of chronic wasting disease in white-tailed deer Ecosphere 4(1).
Thompson A, Samuel M, Van Deelen T. 2008. Alternative feeding strategies and potential disease transmission in Wisconsin White-tailed deer. Journal of Wildlife
Management 72(2):416-421.
Thrusfield M. 2007. Veterinary Epidemiology. 3rd Ed. Blackwell Science, Carlton.
Todd D. 2000. Circoviruses: Immunosuppressive threats to avian species: A review
Avian Pathology 29(5):373-394.
TRAFFIC. 2011. TRAFFIC bulleting seizures and prosecutions. Cambridge.
Travis D, Watson R, Tauer A. 2011. The spread of pathogens through trade in wildlife.
In: The spread of pathogens through international trade in animals and animal products, S. Macdiarmid, (editor.). Rev. sci. tech. Off. int. Epiz, pp. 219-239.
UNEP. 1992. Convention on Biological Diversity. UNEP, Nairobi, Kenya.
Valiente Moro C, De Luna C, Tod A, Guy J, Sparagano O, Zenner L. 2009. The poultry red mite (Dermanyssus gallinae): a potential vector of pathogenic agents.
Experimental & Applied Acarology 48:93-104.
216 Valiente Moro C, Fravolo P, Amelot M, Chauve C, Zenner L, Salvat G. 2007.
Colonization and organ invasion in chicks experimentally infected with Dermanyssus gallinae contaminated by Salmonella Enteritidis. Avian Pathology 36(4):307-311.
Varsani A, Regnard GL, Bragg R, Hitzeroth II, Rybicki EP. 2011 Global genetic diversity and geographical and host-species distribution of beak and feather disease virus isolates. Journal of General Virology 92 (4 ):752-767
Wade L. 2006. Knemidocoptiasis in birds. Veterinary Medicine 101:782-790.
Walberg J. 2001. White blood cell counting techniques in birds. Seminars in Avian and Exotic Pet Medicine 10(2):72-76.
Walker S, Bosch J, James T, Litvintseva A, Valls J, Piña S, García G, Rosa G,
Cunningham A, Hole S, Griffiths R, Fisher M. 2008. Invasive pathogens threaten species recovery programs. Current Biology 18(18):853-854.
Ward J, Gartrell B, Conklin J, Battley P. 2011. Midazolam as an adjunctive therapy for capture myopathy in bar-tailed godwits (Limosa lapponica baueri) with prognostic indicators. Journal of Wildlife Diseases 47(4):925-935.
Warnecke L, Turner J, Bollinger T, Lorch J, Misra V, Cryan P, Wibbelt G, Blehert D,
Willis C. 2012. Inoculation of bats with European Geomyces destructans supports the novel pathogen hypothesis for the origin of white-nose syndrome. Proceedings of the National Academy of Sciences 109(18):6999-7003.
West G, Heard D, Caulkett N. 2007. Zoo Animal and Wildlife Immobilization and
Anesthesia. Blackwell Publishing Co.
217 White D, Hall R, Jakob-Hoff R, Wang J, Jackson B, Tompkins D. 2014. Exudative cloacitis in the kakapo (Strigops habroptilus) potentially linked to Escherichia coli infection. New Zealand Veterinary Journal.
Whiteman NK, Sanchez P, Merkel J, Klompen H, Parker PGD-. 2006. Cryptic host specificity of an avian skin mite (Epidermoptidae) vectored by louseflies
(Hippoboscidae) associated with two endemic Galapagos bird species. The Journal of
Parasitology 92(6):1218-1228.
Williams E, Thorne E. 1996a. Exertional myopathy.In: Non-infectious diseases of wildlife, A. Fairbrother, L. Locke, AND G. Hoff, (editors.). Iowa State Press, Ames,
Iowa.
Williams E, Thorne E. 1996b. Exertional myopathy (capture myopathy). 2nd Ed.
Manson Publishing Ltd, London, UK.
Wobeser G, Wobeser A. 1992. Carcass disappearance and estimation of mortality in a simulated die-off of small birds. Journal of Wildlife Diseases 28(4):548-554.
Wright D, Clout M. 2001. The eastern rosella (Platycercus eximius) in New Zealand.
Department of Conservation, Wellington.
Yunker CE. 1955. Apparent extrinsic variation in Knemidokoptes pilae Lavoipierre and
Griffiths, 1951 (Acarina:Sarcoptiformes). The Journal of Parasitology 41(6):642-643.
Yunker CE, Ishak KG. 1957. Histopathological observations on the sequence of infection in knemidokoptic mange of budgerigars (Melopsittacus undulatus). The
Journal of Parasitology 43(6):664-672.
218
219 APPENDIX 1: Data sheet for field captures
DATA SHEET kakariki project
Date ID Lat Long Band L: R: Recap (y/n)
Timings Capture time: Release time: Extract. Start: Extract. End: Anaesth. Start: Anaesth. End: Recovery:
Physical data Species Cyanoramphus novaezelandie Handler Bird WT Sex Condition Age (Ad/Juv)
Culmen width mm Wing chord mm
Culmen length mm Tarso-metatarsus
Anaesthetic data Temp Fluids given? HR RR
Sample In? With? Amount Tick when done
Blood in lysis buffer Clear tube Longmire's buffer 2 drops blood Blood on filter paper (2) Filter paper N/a 2 x 10mm circle Blood for biochem Yellow top tube N/a 0.4ml
Blood in haematocrit Micro tube N/a Fill 2/3rds of tube Blood smears Slide N/a 2 slides
Faeces for float Flip top 1.5ml tube N/a Large portion
Faeces for freezing Flip top 1.5ml tube N/a Small portion Skin biopsy Formalin N/a 3-5mm biopsy Feather in ethanol 5ml pot 70% ethanol 2-4 blood feathers Feather in envelope Envelope N/a 2 blood feathers Feather in formalin 1.5ml screw top10% formalin 2 blood feathers
Skin for parasitology Colour fliptop tube ethanol 70% 2mm biopsy :
Parasites in ethanol Colour fliptop tube ethanol 70% all found s
e t
Photos taken Bird banded All samples taken o N
220 DATA SHEET kakariki project
Physical exam Body condition Poor Average Good Very good Comments
Skin/feathers Stress bars Feather loss pin feathers dystrophic feathers Comments
Eye exam Normal Abnormal Comments
Oral exam Normal Abnormal Comments
Abdo palp Normal Abnormal Comments
Chest auscultation Normal Abnormal Comments
Cloacal exam Normal Abnormal Comments
General comments
221