INTRODUCTION
The endangered yellow-eyed penguin, an endemic species of southern New Zealand, has suffered major population declines and periodic mass mortality. These events involved losses of up to 60% of the breeding adults in some areas, and may have a significant impact on the survival of the species (Darby and Seddon, Reference Darby, Seddon, Davis and Darby1990). The cause of these mortality events, and the possible role of infectious disease, has never been conclusively determined (Duignan, Reference Duignan2001; McKinley, Reference McKinley2001; Darby, Reference Darby2003; Alley et al. Reference Alley, Morgan, Gill and Hocken2004). Previous declines have been attributed to avian malaria, food shortage and marine biotoxins, but without confirmation of the suspect agent (Darby and Seddon, Reference Darby, Seddon, Davis and Darby1990; Van Heezik, Reference Van Heezik1990; Moore et al. Reference Moore, Murray, Mills, McKinlay, Nelson, Murphy and Moore1991).
The total estimated breeding population of yellow-eyed penguins is approximately 2000 pairs (McKinley, Reference McKinley2001). They nest in sparse colonies and are dependent on their limited terrestrial habitat to breed (Darby and Seddon, Reference Darby, Seddon, Davis and Darby1990). The main nesting areas are Stewart Island (~170–320 pairs), Codfish Island (50–80 pairs) (Darby, Reference Darby2003), the subantarctic Auckland and Campbell Islands (1010–1170 pairs) and the South Island (300–320 pairs) (McKinley, Reference McKinley2001). The South Island has 3 main coastal colonies at Oamaru, the Otago Peninsula and the Catlins.
Stewart Island is the largest nesting region outside the subantarctic, but often features high chick mortality and low reproductive success compared to other nesting sites (King, Reference King2007). Examination of chick mortalities in 2005 on Stewart and Codfish Islands identified 8 cases of leucocytozoonosis, a previously unreported disease of yellow-eyed penguins (Alley, Reference Alley2005). Other diseases identified in yellow-eyed penguins include aspergillosis, malaria (Plasmodium), and the syndrome of diphtheritic stomatitis, which caused major chick mortality in the Otago coast in 2002 (Alley et al. Reference Alley, Morgan, Gill and Hocken2004). A single case of mortality due to Plasmodium, a related Apicomplexa to Leucocytozoon spp., was diagnosed in a yellow-eyed penguin from the Otago peninsula in 2004 (Alley et al. Reference Alley, Morgan, Gill and Hocken2004). However, a recent study failed to identify Plasmodium spp. in 143 yellow-eyed penguins from Otago Peninsula (Sturrock and Tompkins, 2007), despite historical studies finding high seroprevalence (Graczyk et al. Reference Graczyk, Cockrem, Cranfield, Darby and Moore1995 a, Reference Graczyk, Cranfield, Brossy, Cockrem, Jouventin and Seddonb).
Leucocytozoon is a haemoparasite of the Apicomplexa phylum that is considered to be host-specific at the family level (Peirce et al. Reference Peirce, Greenwood and Stidworthy2005) and is found in birds worldwide. Leucocytozoon has been identified in Fiordland crested penguins (Eudyptes pachyrhynchus) in New Zealand (Fallis et al. Reference Fallis, Bisset and Allison1976), whose range overlaps that of yellow-eyed penguins on Stewart Island, in captive Macaroni (Eudyptes chrysolophus) (Peirce et al. Reference Peirce, Greenwood and Stidworthy2005) and in African (Spheniscus demersus) penguins (Earle et al. Reference Earle, Huchzermeyer, Bennet and Brossy1993). On the southern islands of New Zealand, the sporogony phase of Leucocytozoon occurs in simulid flies, particularly Austrosimulium ungulatum, the only fly observed feeding on Fiordland crested penguins (Allison et al. Reference Allison, Desser and Whitten1978). Sporogonic stages were also observed in the midgut and salivary gland of collected flies from ookinetes to mature oocysts containing sporozoites (Desser, 1967). Similar to simuliids carrying L. simondi from ducks, those flies feeding on lightly infected chicks during the September to November period had longer survival times than those feeding on heavily infected individuals (Allison et al. Reference Allison, Desser and Whitten1978).
Leucocytozoon affects circulating leucocytes and erythrocytes as well as tissue macrophages and endothelial cells, where in the latter it creates large tissue schizonts up to 700 μm in diameter (Atkinson and Van Riper III, Reference Atkinson, Van Riper, Loye and Zuk1991). Some Leucocytozoon spp. are pathogenic, causing mortality or reduced fertility and growth. Epizootic infections with L. simondi have caused seasonal mass-mortality of Canada swans (Fallis et al. Reference Fallis, Desser, Khan and Dawes1974; Herman et al. Reference Herman, Barrow and Tarshis1975; Atkinson and Van Riper, III Reference Atkinson, Van Riper, Loye and Zuk1991), while L. caulleryi has threatened waterfowl (Fallis and Desser, Reference Fallis, Desser and Kreier1977), and caused mortality and decreased production in poultry across Asia (Morii, Reference Morii1992). In addition, Merino et al. (Reference Merino, Moreno, Sanz and Arriero2000) demonstrated detrimental effects of Leucocytozoon on host reproductive success and body condition in blue tits (Parus caeruleus).
Thus, with the previous identification of Leucocytozoon spp. and the inconsistent evidence of Plasmodium spp. infection in the yellow-eyed penguin population, the aim of this paper was to further investigate whether Leucocytozoon, is involved in the observed chick mortalities on Stewart Island, and establish the prevalence of infection in the wider population of yellow-eyed penguins in New Zealand.
MATERIALS AND METHODS
This study involved yellow-eyed penguins identified at routine nest checks on the Otago coast of the South Island and on Stewart Island during the November 2006 to January 2007 breeding season (Fig. 1). Penguins on Stewart Island were monitored along the north-eastern Anglem coast (46°48′S 168°01′E to 46°41′S 167°47′E) while those penguins located in the South Island were monitored from Oamaru (45°05′S 170°59′E), Otago Peninsula (45°46S 170°42′E) and the Catlins (46°27′S 169°49′E). The samples collected for this study were collected as an adjunct to the ongoing monitoring of nest sites. All penguins were caught by hand at nesting sites. As a result only breeding adults and chicks below the age of fledging (~90 days) were sampled.
Major yellow-eyed penguin nesting sites on the South Island and Stewart Island used in this study. Penguins on Stewart Island were monitored along the north-eastern Anglem coast (46°48′S 168°01′E to 46°41′S 167°47′E) while those penguins located in the South Island were monitored from Oamaru (45°05′S 170°59′E), Otago Peninsula (45°46S 170°42′E) and the Catlins (46°27′S 169°49′E).
Blood collection and smear preparation
Blood samples were collected using the ulnar or brachial vein from a total of 107 yellow-eyed penguins, from the South Island (Catlins (n=30), Otago Peninsula (n=33) and Oamaru (n=32)), and Stewart Island (n=12). Blood samples were stored in lithium heparin vials and frozen at −20°C.
Thin blood smears were prepared at the time of collection from a total of 49 penguins, from the South Island Catlins region (n=17 adults and 18 chicks) and Stewart Island (n=12 adults and 2 chicks). There is a discrepancy between the number of blood samples and the number of thin smears taken because of the remote nature of the field operations, the inclement weather conditions and the poor quality of many blood smears obtained. The blood smears were air dried, fixed in absolute methanol and subsequently stained with Wright's stain. Each smear was examined under a light microscope for 10 min, initially at 400×, then at 1000×.
Clinical and pathological studies
Twenty-five of the 32 chicks which died on Stewart Island during the 2006/07 season and were of suitable condition underwent post-mortem examination using standard methods (Rae, Reference Rae, Harrison and Lightfoot2006). Samples of liver, spleen, lung, kidney and brain were frozen or placed in 10% buffered formalin, depending on its availability in the field. Histopathlogy was performed on tissue from 14 chicks where suitable fixed tissues were available. The tissues were routinely processed, embedded in paraffin, cut at 3 μm and stained with haematoxylin and eosin.
Molecular studies
DNA was extracted from frozen heparinized whole blood obtained from 107 adults or frozen liver tissues from 7 chicks using a Qiagen DNeasy Kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions for nucleated whole blood or tissue DNA extraction where appropriate. The presence of the cytochrome b gene of Leucocytozoon was identified using a nested PCR and the primer sets HaemNF1/HaemNR3 and HaemFL/HaemR2L as described by Hellgren et al. (Reference Hellgren, Waldenstrom and Bensch2004). To confirm successful amplification, 10 μl of the final PCR product was run on a 1·5% agarose gel containing ethidium bromide, prior to purification and sequencing.
All Leucocytozoon-positive PCR amplicon samples were purified using a PureLink PCR purification kit (Invitrogen, Carlsbad, CA, USA) and subjected to automatic dye-terminator cycle sequencing with BigDyeTM Terminator Version 3.1 Ready Reaction Cycle Sequencing kit and the ABI3730 Genetic Analyzer (Applied Biosystems Inc, Foster City, CA, USA) to confirm the genomic sequence.
Analysis of the Leucocytozoon sequences (approximately 416 bases) obtained from yellow-eyed penguins was compared by NCBI Blast to other published sequences available from GenBank. Isolates and known GenBank sequences were trimmed to the same length (402 bases) using GeneiousTM (Biomatters, Auckland, New Zealand) and aligned using Clustal W (Higgins et al. Reference Higgins, Thompson, Gibson, Thompson, Higgins and Gibson1994) with gaps ignored. The resulting sequences were submitted to the GenBank database (GU065716, GU065717, and GU065717). A phylogenetic tree was generated using a Jukes-Cantor distance model and the Neighbour-joining method using the trimmed Clustal W alignment and performed in Geneious™. Bootstrap values were generated from 1000 cycles. The sequence divergence between and within the different lineages was calculated using a Jukes-Cantor model of substitution implemented in the program PAUP* 4.0 Beta version 10 (Swofford, Reference Swofford2002).
Statistics
The small sample size limited statistical analysis of the results. Freecalc 2 (Cameron, Reference Cameron2001), Win Episcope 2.0 (Blas et al. Reference Blas, Ortega, Frankena and Noordhuizen2000) and reference tables (Beyer, Reference Beyer and Thrusfield2005) were used to generate 95% confidence intervals for the true prevalence based on a low expected prevalence (3–5%). The range of the true prevalence is reported at the 95% confidence interval.
RESULTS
Blood smears
Examination of 49 blood smears taken from the yellow-eyed penguin adults or chicks from the South Island Catlins region and Stewart Island found no evidence of Leucocytozoon spp. parasitaemias or abnormalities of red cell morphology by light microscopy (Table 1).
Blood smear, PCR, histology and prevalence results for Leucocytozoon spp. from adult (A) and chick (B) yellow-eyed penguins on the South Island and Stewart Island during the study period
# Results given as positive/total sampled.
† 95% level of confidence.
* Smears collected for penguins >10 days old.
Clinical and pathological findings
All 32 monitored chicks on Stewart Island died during the 2006/07 breeding season. They were pale, grossly underweight at 100–200 g with poor feather growth, and exhibited lethargy, poor body condition (BCS <2/9) and muscle atrophy. Seven chicks (22%) developed swollen eyes and peripheral oedema. Open-mouth breathing and respiratory distress were often noted just prior to death. Deaths occurred most commonly at 10 days old (mean 13·8, standard deviation 18·4), with a range of 2–108 days, as shown in Fig. 2. Chicks died within 12–72 h after the onset of clinical signs.
Distribution of age at death for chicks on Stewart Island from November 2006 to February 2007.
Post-mortem findings from 25 chicks which died on Stewart Island included generalized pallor (n=22, 88%), splenomegaly (n=4, 16%) and hepatomegaly (n=5, 20%). Variable ecchymotic haemorrhages of the pericardial, serosal and subcutaneous surfaces (n=3, 12%), and pulmonary haemorrhage (n=2, 8%) were seen in severe cases. The stomach was full in 7 chicks, containing fish material and/or nesting material, empty in 6 others, and unreported in the remaining 12. The oldest chick that died at 108 days of age exhibited poor body condition, pale musculature, extensive subcutaneous, serosal and pericardial ecchymotic haemorrhages, splenomegaly and pulmonary haemorrhage. Of the 14 chicks examined histopathologically, 12 showed no evidence of primary schizont or megaloschizont formation in their tissues; whereas, 2 (14%) chicks showed severe, disseminated megaloschizont formation characteristic of leucocytozoon in the liver, spleen, lung, kidney and other tissues (Fig. 3). These birds were aged 22 and 108 days, and were the oldest chicks sampled.
Photo-micrograph showing a hepatic megaloschizont (M) from a yellow-eyed penguin chick, separated from hepatocytes (H) by a thick capsule (C) (1000×).
Molecular studies
Leucocytozoon DNA was amplified from 10 of the 119 (8%) yellow-eyed penguin adult blood samples as shown in Table 1, reflecting infection in 10 of 12 (83%) blood samples taken from the adults on Stewart Island. In addition, frozen liver samples from 7 chicks (aged 9 to 108 days) from Stewart Island examined post-mortem were positive for Leucocytozoon DNA, and in 1 of these the presence of megaloschizonts in tissues was confirmed histopathologically. Leucocytozoon spp. DNA was not detected by PCR in the blood samples of adult South Island yellow-eyed penguins from the Catlins (n=30), Otago Peninsula (n=33) or Oamaru (n=32) regions. Resulting sequences were used for BLAST analysis to compare the yellow-eyed penguin Leucocytozoon isolates against the NCBI GenBank database. The results showed that all but 1 isolate had 99% sequence homology with known Leucocytozoon spp. isolated from a barn owl (Tyto alba, GenBank EU627792), spotted owl (Strix occidentalis, GenBank EU627793, and a marsh harrier (Circus aerginosus, GenBank EF607287). The remaining isolate from a yellow-eyed penguin chick (no. 7) showed high sequence homology to the same 3 database sequences, but at a slightly lower percentage, 98%, 97%, and 97% respectively, suggesting slight sequence variation in this isolate compared to the other yellow-eyed penguin Leucocytozoon isolates.
Statistical analysis of PCR results
The test prevalence of Leucocytozoon in the yellow-eyed penguin population examined was 8·4% with the true prevalence of Leucocytozoon falling between 4·5–16% at the 95% level of confidence. The test prevalence of Leucocytozoon on Stewart Island was 83% with the true prevalence of infection in this population falling between 52 and 98% at the 95% level of confidence. If we exclude the Stewart Island population, our level of sampling meant that the maximum possible prevalence of Leucocytozoon that was not detected in the South Island yellow-eyed penguin population was up to 2·8%.
Sequencing
A multiple alignment was performed using Clustal W on 402 bp of the cytochrome B gene of Leucocytozoon isolated from the blood or tissues from the Stewart Island yellow-eyed penguin population. A sequence divergence and phylogenetic tree constructed using the Neighbour-joining method resulted in sequences of Leucocytozoon DNA from yellow-eyed penguins forming a distinct cluster within the Leucocytozoon lineage consistent with the morphology of this parasite (Fig. 4). As a result, the isolates have no direct genetic relationship with a sequence divergence of 14·4%, to a Plasmodium relictum sp. isolated from a African penguin (Spheniscus demersus, GenBank NC012426). Although a distinct subclade of Leucocytozoon spp., the majority of the isolates do associate with other previously published sequences consisting of Leucocytozoon spp. from a common loon (Gavia immer, GenBank EF077166) with a sequence divergence of 2·2%, a barn owl (Tytl alba, GenBank EU6227792) with a sequence divergence of 0·5%, a marsh harrier (Circus aeruginosus, GenBank EF607287) and spotted owl (Strix occidentalis, GenBank EU627793) with a sequence divergence of 0·7% (Fig. 4). Within the sequences obtained from the yellow-eyed penguin adults and chicks, there was some minor sequence variation, particularly between chick 7 which had a sequence divergence of 1·5% when compared to the other yellow-eyed penguin isolates and approximately 2% sequence divergence from the barn owl, spotted owl and marsh harrier (Fig. 4). These results are consistent with the BLAST results and suggestive of a small level of sequence divergence within the yellow-eyed penguin population on Stewart Island.
Phylogenetic analysis of Leucocytozoon spp. isolated from yellow-eyed penguins. Neighbour-joining (NJ) phylogeny of mitochondrial cytochrome b gene from 2 lineages of Plasmodium spp., 11 Leucocytozoon sp. submitted to GenBank, and Leucocytozoon isolated from yellow-eyed penguins (adult n=10 and chick n=7) living on Stewart Island. The tree is rooted on a lineage of Haemoproteus spp. Numbers above the branches indicate bootstrap support based on 1000 replicates. Names of the lineages (when available) and GenBank Accession numbers of the sequences are given after the species names of the parasites.
DISCUSSION
This study identifies a prevalence of Leucocytozoon infection in yellow-eyed penguins during the 2006–2007 breeding season and suggests an association with an observed pattern of low reproductive success. A definitive link between Leucocytozoon and nestling deaths was not confirmed although histological examination suggests that older chicks (n=2) may have died with disseminated leucocytozoonosis. Leucocytozoon infection was characterized by acute signs in chicks with megaloschizont formation in tissues and chronic infection in 10 of 12 (83%) clinically normal adults detected by blood PCR which was not observed in any of the blood smears examined, suggestive of ongoing schizogony (Fallis et al. Reference Fallis, Desser, Khan and Dawes1974), subclinical infection or recrudescence of latent infection (Cranfield et al. Reference Cranfield, Graczyk, Beall, Ialeggio, Shaw and Skjoldager1994; Graczyk et al. Reference Graczyk, Cranfield, Brossy, Cockrem, Jouventin and Seddon1995 b; Brossy et al. Reference Brossy, Plos, Blackbeard and Kline1999).
These results showed that Leucocytozoon infection was far more prevalent on this island than previously realized. In addition, infection was limited exclusively to Stewart Island, with no positive results found at any of the other locations on the South Island. The distribution in this study correlates with low reproductive success on Stewart Island, which has been between 21 and 33% since 2003, compared to the South Island Catlins region (60–99%) and neighbouring Codfish Island (48–57%) over the same period (King, Reference King2007). Interestingly, breeding success in the Bravo Islands, a small cluster of islands within the Stewart Island group, more closely resembled that on Codfish Island rather than Stewart Island (King, Reference King2007). The reasons for this distinct variation in breeding success and Leucocytozoon prevalence are unclear but are suggestive of localized factors on Stewart Island.
Investigation into the causes of chick mortality has been stimulated in part by evidence that predation by cats on Stewart Island has been minimal, whereas predation by cats and mustelids on the South Island, where nesting sites are often adjacent to farmland, tourism or recreational areas, is significant (King, Reference King2007). On Stewart Island, starvation and disease are proposed as key factors in failure to fledge (Darby and Seddon, Reference Darby, Seddon, Davis and Darby1990; Massaro and Blair, Reference Massaro and Blair2003; King, Reference King2007).
Starvation may have been a contributing factor to chick mortalities. It is possible that there was sufficient nutrition to maintain condition in adults, but not rapidly growing chicks. Starvation may be due to poor delivery, acceptance, digestion or quality of food. Evidence of recent feeding activity in chicks, as indicated by stomach contents at post-mortem examination suggests that active food delivery occurs in many unwell chicks, suggesting that starvation may be a secondary factor.
In yellow-eyed penguin chicks, rapid growth and moulting confer a considerable degree of physiological stress. Furthermore, excessive immune response to pathogens may directly cause inappetance and direct nutrients away from homeostasis and production (Deitemeyer, Reference Deitemeyer and Deitemeyer2005). Allison et al. (Reference Allison, Desser and Whitten1978) also suggested that heavy Leucocytozoon infection may render chicks more susceptible to secondary infection or stress. In some avian species, heavy infections, or infections involving multiple haematozoa species, are capable of causing severe anaemia, tissue inflammation and death (Herman et al. Reference Herman, Barrow and Tarshis1975; Evans and Otter, Reference Evans and Otter1998), while evidence in other species has suggested the that true impact lies in decreased reproductive performance (Merino et al. Reference Merino, Moreno, Sanz and Arriero2000; Dunbar et al. Reference Dunbar, Torniquist and Giordano2003).
Yellow-eyed penguin chicks are exposed to large numbers of biting black flies from hatching and are not protected in the nest by parents as observed in Fiordland crested penguins (Allison et al. Reference Allison, Desser and Whitten1978). Early inoculation combined with short haemoparasite pre-patent periods, which have been recorded as early as 4 days in some Leucocytozoon spp., could account for early chick deaths due to leucocytozoonosis (Desser, Reference Desser1967; Khan and Fallis, Reference Khan and Fallis1970; Fallis et al. Reference Fallis, Desser, Khan and Dawes1974). The suspected vector, Austrosimulium ungulatum, increases feeding activity with warmer temperatures, peaking in early morning and late afternoon over summer, coinciding with chick rearing (Allison et al. Reference Allison, Desser and Whitten1978). Occasional conjunctival petechial haemorrhages observed in nesting adults on Stewart Island may be caused by biting vectors (unpublished observations). Thus, adults may be exposed to low levels of re-infection throughout the year, stimulating immunity, while chicks are exposed to high levels of Leucocytozoon during the peak transmission period of the breeding season (Allison et al. Reference Allison, Desser and Whitten1978). The overwhelming presence of simuliid vectors on Stewart Island is suggestive of its role as a vector; however, conclusive evidence would require transmission studies. Further circumstantial evidence for the role of suspected vectors could be gathered by conducting histological examination or PCR surveys of the simuliids for evidence of Leucocytozoon infection.
The result of this study found that PCR was superior to peripheral blood smears in the detection of Leucocytozoon from clinically normal birds. This finding is consistent with other studies surveying low-level parasitaemias (Jarvi et al. Reference Jarvi, Schultz and Atkinson2002; Richard et al. Reference Richard, Sehgal, Jones and Smith2002; Swinnerton et al. Reference Swinnerton, Pierce, Greenwood, Chapman and Jones2005). Jarvi et al. (Reference Jarvi, Schultz and Atkinson2002) found that the sensitivity (Bayes' theorem) of detecting the chronic phase of Plasmodium by microscopy was 0·27 (0–0·67) compared to 0·61 for PCR (0·2–0·87). Hellgren et al. (Reference Hellgren, Waldenstrom and Bensch2004) showed 100% positive amplification from circulating Leucocytozoon levels over 1 gametocyte per 10 000 blood cells, but only 67% with levels of 1 per 100 000 blood cells for the primers used in this study. The short window when life stages of Leucocytozoon are present in peripheral blood made detection by smears difficult and may also cause underestimation of prevalence determined by nested PCR of blood samples (Jarvi et al. Reference Jarvi, Schultz and Atkinson2002; Hellgren et al. Reference Hellgren, Waldenstrom and Bensch2004; Swinnerton et al. Reference Swinnerton, Pierce, Greenwood, Chapman and Jones2005).
The presence of highly similar Leucocytozoon sequences in adults and chicks indicates that adults may serve as a reservoir host for the infection of chicks. In addition, the presence of a slightly different (1·5%) Leucocytozoon sequence in 1 chick (no. 7), suggests that there may either be several circulating Leucocytozoon species, including minor species, in the yellow-eyed penguin population or this variant was introduced as a result of interactions with other avian hosts in the presence of the vector. The penguin host maintains infection during the non-breeding season, so the introduction of dispersing, infected penguins in a naïve population with appropriate vectors could have severe and widespread effects. Unfortunately, morphological identification was not available for infections detected in the adult birds due to the absence or low level of circulating gametocytes.
The similarity of the Leucocytozoon molecular sequences to those recovered from raptors and waterfowl suggests that other wild species present on Stewart Island, such as Australasian harriers (Circus approximans) and ducks (Anas spp.), may be the origin of the novel Leucocytozoon described in this study. Hellgren et al. (Reference Hellgren, Waldenstrom, Perez-Tris, Szollosi, Hasselquist, Krizanauskiene, Ottosson and Bensch2007) suggest that Leucocytozoon are restricted to one resident bird fauna over a long evolutionary time-span. More investigation is required, particularly of the molecular phylogeny of the Leucocytozoon variants present in this ecosystem, including simuliid vectors and alternate hosts before the evolutionary origins of the haemoparasite in yellow-eyed penguins can be determined. Leucocytozoon is considered host specific at the family level, and Allison et al. (Reference Allison, Desser and Whitten1978) demonstrated vector transfer of L. tawaki from a Fiordland crested to a juvenile blue penguin (Eudyptula minor), so it is possible that the close nesting proximity of these species and high vector density on Stewart Island could allow interspecies spread.
Although we have documented leucocytozoonosis in yellow-eyed penguins, it remains unclear what role Leucocytozoon infection plays in reproductive success and mortality. The clinical signs exhibited by chicks are consistent but non-specific, with poor growth and development being a common finding in a number of neonatal conditions. L. simondi and L. smithi both cause lethargy, inappetance, dyspnoea, anaemia and death and are associated with splenomegaly and hepatomegaly as seen in yellow-eyed penguins (Fallis et al. Reference Fallis, Desser, Khan and Dawes1974). Infection with L. smithi in turkeys (Meleagris gallopavo) produced respiratory distress caused by alveolar capillary blockage, but this finding was not observed histologically in penguins (Siccardi et al. Reference Siccardi, Rutherford and Derieux1974). Alternatively, the physiological, behavioural or ecological costs of the Leucocytozoon may have combined with other factors such as poor food supply, climatic extremes or concurrent infections with other pathogens. It is likely that the expression of mortality due to haemoparasites is a multifactorial interaction between the host, the pathogen and the environment (Galvani, Reference Galvani2003).
The Leucocytozoon species found in this study exhibited a large megaloschizont tissue phase, which may be a factor in its association with morbidity and mortality. Fallis et al. (Reference Fallis, Desser, Khan and Dawes1974) noted that pathogenic species, including L. simondi, L. caulleryi and L. sakharoffi, feature megaloschizont formation while those species that cause less severe disease do not. The observed anaemia in acute infections is thought to be caused by anti-erythrocytic factors released at the time of schizont rupture, although confirmation of this was outside the scope of this study (Atkinson and Van Riper III, Reference Atkinson, Van Riper, Loye and Zuk1991). Further investigation is warranted to quantify and characterize anaemia to determine its role in penguin mortalities.
The impact of ongoing chick losses on the survival of the yellow-eyed penguin population is ameliorated by the longevity of adults, but continued chick losses of the magnitude observed will result in a high extinction probability for these populations. The Stewart Island and Codfish Island breeding group represent a significant population for the species and dispersing juveniles may travel to both the South Island and the subantarctic islands (Department of Conservation, 1991). Stewart Island yellow-eyed penguins may also represent a genetically distinct group which is highly important in an endangered species with a limited breeding pool (Triggs and Darby, Reference Triggs and Darby1989). Therefore, wider studies are needed including Stewart Island and the subantarctic islands, which constitute the bulk of the yellow-eyed penguin population, and of other birds on these islands which may represent alternate hosts. It may also be useful to determine the incidence of Leucocytozoon infection over multiple years to reliably estimate its effect on the population dynamics of the yellow-eyed penguin.
