Abstract
The Bali mynah (Leucopsar rothschildi) is an endangered bird endemic to the island of Bali, Indonesia. The wild population is estimated to be approximately nine (T. Norton, personal communication), while the North American Species Survival Plan (SSP) population consists of 232 individuals distributed among 62 institutions as of September 2000.11 Atoxoplasmosis and hemochromatosis are considered to be the primary medical problems in the captive population.9 Although a great deal of work has been done on describing the agent, clinical signs, diagnostic methodology and pathology of atoxoplasmosis, no American Zoo and Aquarium Association (AZA) wide effort has been made to identify and assess potential risk factors associated with captive management of the Bali mynah. This study was designed to survey the AZA facilities holding Bali mynahs in order to derive an institutional prevalence for both atoxoplasmosis and hemochromatosis, and to elicit management-related information to identify risk factors for atoxoplasmosis in the Bali mynah.
Introduction
Atoxoplasmosis, the actual disease process caused by Atoxoplasma spp., recently classified as Isospora rothschildi, is a coccidian parasite of passerine birds with a prolonged life cycle that involves the reticuloendothelial system as well as the intestinal epithelium.10 Clinical signs associated with atoxoplasmosis are often nonspecific and may include diarrhea, decreased appetite, weight loss and ruffled feathers;10 however, infestation of fledgling birds between 3 and 8 wk may result in acute death7. Hemochromatosis (iron storage disease) is caused by excessive iron accumulation in numerous organs throughout the body, especially the liver. This process commonly leads to end-stage liver disease in the Bali mynah as well as other birds. Twenty-one percent (15/70) of postmortems performed by the Bali mynah SSP pathology advisor showed some degree of this disease.8 A complete review of these diseases, including clinical signs, diagnostic protocols and methodology is available from the SSP veterinary advisor.8
Atoxoplasmosis was first reported in North American Bali mynahs at the National Zoo in 1989.7 Retrospective studies showed that the organism had been present in Bali mynahs at the National Zoo since 1975.10 Atoxoplasmosis has been described previously in numerous other species of birds (sparrows, canaries, evening grosbeaks, rose breasted grosbeaks, common mynahs, Indian hill mynahs).7 Several reports exist in the literature regarding atoxoplasmosis in Bali mynahs in North American institutions. One report mentioned that 8/9 (89%) U.S. institutions screened were positive,7 while Greiner et al. found at least one positive Bali mynah in 34/36 (84%) facilities and a minimum prevalence of 62% (152/245 examined) based upon the presence of oocysts in the feces and/or zoites in mononuclear phagocytes.5 A study of atoxoplasmosis in captive Bali mynahs in the U.K. showed an 18.5% bird prevalence based on fecal smears and an 47% (9/19) institutional prevalence.3 Though oocysts have been found in the feces of wild Bali mynahs, it is not yet known the extent to which atoxoplasmosis has contributed to the wild population’s decline.9,2
Methods
A survey designed to retrospectively collect both medical and husbandry information was distributed to 55 AZA-accredited zoologic institutions currently or previously holding Bali mynahs. Veterinarians and animal management staff were asked to complete the survey either on paper or on the 31/2" floppy disc provided and return it to the Lincoln Park Zoo in Chicago. Upon receipt, completed surveys were coded in order to ensure confidentiality and data entered into an Access database. Descriptive data were summarized using Microsoft Excel spreadsheet software and statistics (odds ratios [OR], chi-square and confidence intervals) were calculated using the Center for Disease Control’s Epi Info 2000 software.4
After descriptive summation of results, institutions were labeled as a “case” if they had diagnosed Atoxoplasma spp. in their collection at any time. Chi-square tests (Yates corrected chi-square) for significance were performed on each relevant question in the survey. In addition, extended Mantel-Haenszel chi-square for trend analysis was performed where appropriate. For all chi-square tests, p values of less than 5% (p<0.05) were considered significant. Odds ratios comparing cases (case vs. non-case) to specific exposure variables (exposure vs. no exposure) were calculated using 2x2 tables. The odds ratio is an indicator of the degree of association between cases and exposure to a specific variable. Odds ratios less than one represent a decreased relationship between cases and exposure and may be considered protective if biologically plausible; an odds ratio of one means that no relationship exists between cases and exposure; while odds ratios greater than one represent an increased likelihood that exposure increases risk of being a ‘case.’ Ninety-five percent confidence intervals (Cornfield approximation and exact) surrounding odds ratios were calculated; confidence intervals that included ‘one’ were considered not significant.4
Results
Forty out of 55 surveys were returned for a 73% response rate. Of those that responded, 18 (45%) indicated that they had diagnosed Atoxoplasma spp. in at least one Bali mynah. These respondents were classified as cases for the risk factor analysis. Fecal exam (55%) was the most common method used to diagnose Atoxoplasma spp., followed by examination of buffy coat smears (35%) and necropsy (25%). Atoxoplasma spp. was reported in other avian species by 15% of respondents; the yellow-bellied laughing thrush, tanager, golden breasted starling (suspect), superb starling, yellow-billed cardinal and warbler were the other positive species listed. Eighty-five percent of the respondents have treated Bali mynahs for atoxoplasmosis, either after diagnosis or for prevention. Only 10 (10%) responded that they performed routine medical exams, once per year or greater; frequency of procedures within those 10 can be found in Table 1. Sixteen institutions (40%) responded that they had diagnosed hemochromatosis in at least one Bali mynah. Necropsy was by far the most common way that hemochromatosis was diagnosed (40%) followed by liver biopsy (20%). One-third (5/16) of the institutions that reported positive diagnoses have attempted treatment. Almost all (90%) respondents submit adults for necropsy, while 72% submit chicks and 10% responded that they do not submit Bali mynahs for necropsy. When comparing cases (respondents which have diagnosed Atoxoplasma spp. in at least one bird) to non-cases (respondents which have never diagnosed Atoxoplasma ssp. in Bali mynahs), all tests of significance were not statistically significant at the 95% confidence level. Raw data, percent of institutions for each response, odds ratios and 95% confidence intervals for all relevant questions are provided in Table 1.
Table 1. Summary of survey results, calculated odds ratios for risk factors associated with Atoxoplasma spp.-positive institutions and 95% confidence intervals.
|
Question
|
Category choices
|
# Responses
|
%
|
OR
|
CI 95%
|
|
Have ever diagnosed atoxoplasmosis
|
Yes
|
18
|
45
|
|
|
|
No
|
22
|
55
|
|
|
|
Number of BM enclosures
|
1
|
21
|
52.5
|
0.36
|
0.08–1.57
|
|
2
|
9
|
22.5
|
1
|
0.16–5.5
|
|
3
|
4
|
10
|
6
|
0.5–309
|
|
4
|
2
|
5
|
0.6
|
0.01–12.4
|
|
>4
|
4
|
10
|
4.2
|
0.3–230
|
|
Location of enclosures
|
Indoor
|
33
|
82.5
|
2.4
|
0.3–27
|
|
Outdoors
|
11
|
27.5
|
1
|
0.2–5
|
|
Combo in/out
|
3
|
7.5
|
0.6
|
0.01–13
|
|
Enclosure composition (materials used)
|
Concrete
|
38
|
95
|
1.21
|
0.25–6
|
|
Glass
|
16
|
40
|
0.6
|
0.13–2.6
|
|
Wood
|
14
|
35
|
0.56
|
0.11–2.5
|
|
Wire fence
|
30
|
75
|
1.31
|
0.25–7.6
|
|
Plastic sheeting
|
3
|
7.5
|
0.37
|
0.01–5.2
|
|
Polypropylene mesh
|
3
|
7.5
|
0.59
|
0.01–12.4
|
|
Other
|
11
|
27.5
|
1.7
|
0.34–8.8
|
|
Enclosure substrates used
|
Soil/leaf litter
|
33
|
82.5
|
6.38
|
0.63–310
|
|
Gravel
|
9
|
22.5
|
3.17
|
0.53–22
|
|
Concrete
|
14
|
35
|
1.36
|
0.31–6.1
|
|
Other
|
13
|
32.5
|
0.67
|
0.14–3.4
|
|
Enclosure lighting
|
Natural - open (full)
|
18
|
45
|
0.64
|
0.15–2.7
|
|
Natural - skylight (partial)
|
21
|
52.5
|
1.9
|
0.45–8
|
|
Artificial - full
|
7
|
17.5
|
0.9
|
0.11–6.3
|
|
Artificial - partial
|
18
|
45
|
1.4
|
0.34–6.1
|
|
Enclosure cleaning
|
Daily
|
28
|
70
|
0.93
|
0.2–4.3
|
|
Weekly
|
7
|
17.5
|
1.81
|
0.26–14
|
|
Regularly but <weekly
|
2
|
5
|
0.6
|
0.01–12.4
|
|
Only when necessary
|
3
|
7.5
|
2.63
|
0.12–162
|
|
Never
|
1
|
2.5
|
1.24
|
0.02–101
|
|
Enclosure disinfection
|
Daily
|
2
|
5
|
1.24
|
0–49
|
|
Weekly
|
11
|
27.5
|
1
|
0.21–5
|
|
Regularly but <weekly
|
10
|
25
|
1.3
|
0.25–6.8
|
|
Only when necessary
|
7
|
17.5
|
0.9
|
0.13–6
|
|
Never
|
10
|
25
|
0.76
|
0.14–4
|
|
Disinfection products
|
Quaternary ammonium
|
5
|
12.5
|
2
|
0.2–26
|
|
Phenolics
|
1
|
2.5
|
1.24
|
0.02–101
|
|
Chlorhexidine (Nolvasan)
|
3
|
7.5
|
2.63
|
0.12–162
|
|
Iodophors (betadine)
|
2
|
5
|
1.24
|
0.02–101
|
|
Bleach
|
20
|
50
|
0.44
|
0.1–1.9
|
|
Other
|
15
|
37.5
|
0.72
|
0.16–3
|
|
Pest control products
|
Physical invertebrate
|
12
|
30
|
2.16
|
0.5–11
|
|
Chem invertebrate
|
21
|
52.5
|
0.83
|
0.2–3.5
|
|
Physical vertebrate
|
23
|
57.5
|
3.1
|
0.7–15
|
|
Chem vertebrate
|
17
|
42.5
|
0.76
|
0.18–3.3
|
|
Other
|
7
|
17.5
|
0.9
|
0.11–6.3
|
Table 1. Summary of survey results, calculated odds ratios for risk factors associated with Atoxoplasma spp.-positive institutions and 95% confidence intervals (continued).
|
Question
|
Category choices
|
# Responses
|
%
|
OR
|
CI 95%
|
|
Enclosure water source
|
Bowl
|
31
|
77.5
|
3.7
|
0.5–41
|
|
Bottle
|
0
|
0
|
none
|
|
|
Mister
|
3
|
7.5
|
2.6
|
0.12–162
|
|
Enclosure item
|
19
|
47.5
|
0.8
|
0.2–3.3
|
|
Other
|
0
|
0
|
none
|
|
|
Mixed species exhibit
|
Currently yes
|
23
|
57.5
|
3.12
|
0.7–15
|
|
No, but previously
|
9
|
22.5
|
0.53
|
0.07–3
|
|
Never
|
8
|
20
|
0.33
|
0.03–2.3
|
|
Reproductive history
|
Had egg laying
|
31
|
77.5
|
1
|
0.18–6.25
|
|
No egg laying
|
9
|
22.5
|
none
|
|
|
Medical management/exam
|
>1/year
|
3
|
7.5
|
0.6
|
0.01–12
|
|
1/year
|
7
|
17.5
|
0.9
|
0.11–6.3
|
|
Illness/shipment only
|
30
|
75
|
1.31
|
0.25–7.6
|
|
Routine medical exams
|
Physical exam
|
10
|
25
|
|
|
|
Body weight
|
10
|
25
|
|
|
|
CBC
|
9
|
22.5
|
|
|
|
Chem panel
|
9
|
22.5
|
|
|
|
Buffy coat
|
7
|
17.5
|
|
|
|
Serology (Chlamydia)
|
2
|
5
|
|
|
|
Serology (paramyxovirus)
|
0
|
0
|
|
|
|
Aspergillus diagnostics
|
1
|
2.5
|
|
|
|
Fecal (smear and float)
|
10
|
25
|
|
|
|
Fecal for atox
|
6
|
15
|
|
|
|
Fecal cytology
|
2
|
5
|
|
|
|
Fecal culture
|
3
|
7.5
|
|
|
|
Radiographs
|
3
|
7.5
|
|
|
|
Atoxoplasmosis diagnostic methods used to detect positives
|
Fecal
|
22
|
55
|
1.57
|
0.37–6.8
|
|
Buffy coat exam
|
14
|
35
|
0.34
|
0.06–1.6
|
|
Biopsy
|
3
|
7.5
|
0.59
|
0.01–12.4
|
|
Necropsy
|
10
|
25
|
0.43
|
0.06–2.4
|
|
Atoxoplasmosis in other species
|
Yes
|
31
|
77.5
|
3.7
|
0.5–41
|
|
No
|
6
|
15
|
0.56
|
0.05–4.6
|
|
Never mixed
|
3
|
7.5
|
|
|
|
Atoxoplasmosis treatment
|
Have treated
|
5
|
12.5
|
1.89
|
0.45–8.17
|
|
Never treated
|
35
|
87.5
|
|
|
|
Have ever diagnosed hemachromatosis
|
Yes
|
16
|
40
|
2.14a
|
0.5–9.5
|
|
No
|
24
|
60
|
|
|
|
Hemochromatosis diagnostic methods used to detect positives
|
Radiographs
|
3
|
7.5
|
|
|
|
Biopsy
|
7
|
17.5
|
|
|
|
Necropsy
|
16
|
40
|
|
|
|
Other
|
3
|
7.5
|
|
|
|
Hemochromatosis treatment
|
Yes
|
5
|
12.5
|
0.79b
|
0.06–7.9
|
|
No
|
35
|
87.5
|
|
|
|
Routine necropsies
|
Adult necropsy
|
36
|
90
|
2.7
|
0.2–150
|
|
Chick necropsy
|
29
|
72.5
|
2.8
|
0.5–19
|
|
No necropsy
|
6
|
15
|
1.2
|
0.02–101
|
aOR calculated for hemochromatosis as risk factor for atoxoplasmosis.
bOR calculated for hemochromatosis treatment as risk factor for atoxoplasmosis.
Discussion
The absence of statistical significance in this study is most likely due to the small sample size (n≤40); this illustrates the need for high response rates needed when surveying small, endangered populations even with relatively prevalent diseases. Even though all associations did not prove to be statistically significant, overall odds ratio estimates for several factors were quite high; therefore, these factors may prove to be biologically significant, should the power of the study be increased. Odds ratios represent the degree to which a specific factor is related to ‘cases’ in the study; in this case, the degree to which the factor is associated with institutions finding Atoxoplasma spp. in at least one Bali mynah. Statistically non-significant odds ratios (i.e., with a 95% confidence interval including 1) that are elevated and consistent with “common sense” should still be discussed. When evaluating odds ratios, a score of one (1) indicates no relationship, a score greater than one (>1) indicates an increased association between cases and the risk factor, while a value less than one (<1) may represent a factor that is protective for atoxoplasmosis. Risk factors were not evaluated for association with hemochromatosis. Odds ratios that make biologic sense when discussing risk factors associated with atoxoplasmosis are discussed below.
In general, ORs increased as number of enclosures at an institution increased (Table 1). This may be due to a greater number of birds at these institutions, a greater effort put forth to diagnose disease in larger institutions, or may be a result of the increased amount of labor required to thoroughly clean and disinfect a greater number of enclosures. The following enclosure factors had increased ORs and may represent a biologically plausible increase in risk: indoor enclosures (2.4); concrete (1.2) and wire fencing (1.3) enclosure materials; soil/leaf litter (6.4), gravel (3.2) and concrete (1.4) enclosure substrates (notice that risk decreases as substrate is more easily cleaned). Combination indoor/outdoor enclosures (0.6), glass (0.6), wood (0.6), plastic sheeting (0.4) and polypropylene mesh (0.6) were less associated with cases and may be protective, but the potential for confounding between enclosure substrate and disinfection practices must be further investigated.
Partial lighting situations (natural or artificial) appear to be more strongly associated with cases than full lighting, though again this may be due to the association between partial lighting and indoor enclosures. There also appears to be a correlation between frequency of cleaning of an enclosure and institution positivity, with increased OR associated with decreased frequency. Most disinfectants, however, were positively associated with cases (probably because almost everyone uses them); with bleach and the “other” category being the only ones that may be protective. These results, as in the lighting data, may be due more to enclosure design (i.e., decreased prevalence with outdoor) than with the factor of concern. Chemical pest control products (invertebrate and vertebrate) were protective (OR<1), while physical methods (invertebrate and vertebrate) were more strongly associated with cases (OR>1). All water sources represented an increased association (risk) for atoxoplasmosis with the exception of the ‘enclosure item’ category, which had an odds ratio close to one; however, this may again be due to an association with an outdoor enclosure type.
Mixed species exhibits (3.7) were of greater risk than those that previously or never were mixed (OR<1). We must be careful not to put too much weight on this factor since no further interaction was investigated, including observed disease in the other species of birds or type of “coccidia” noted on exam. Risks associated with frequency of medical exams made biologic sense; with those that performed exams once per year or more had an OR of less than one, while those performing exams only upon clinical illness or preshipment had an OR of 1.3. There was a positive association in cases where atoxoplasmosis was diagnosed in other species but these data do not indicate disease directionality. There was a decreased risk of institutions being positive associated with treatment for atoxoplasmosis (0.53) and an increased risk with never having treated (1.9). The increased association with having diagnosed hemochromatosis (2.1) may not be valid because we do not know if they were in the same animal or just in the same institution. This holds for the protective association with hemochromatosis treatment (0.79) as well.
Therefore, we feel that these preliminary data, focusing on information gleaned from the intensive survey sent to Bali mynah-holding institutions, provide evidence of the information that can be acquired through active participation in such surveys, and will form the basis for additional epidemiologic investigations focused on this disease of concern in a critically endangered bird species. While much of the data may not be statistically significant, the directionality of results (indicating a trend towards being a risk factor, a protective factor or unimportant in the overall disease state) will assist in the ongoing allocation of effort (and funds) to better understand this disease state. Only through observing and collecting this type of information on a population-wide basis can these types of questions be addressed in a scientifically defensible manner; therefore, ongoing participation of zoologic veterinarians and animal care staff are critical to the success of such efforts.
Acknowledgments
This work was supported by a grant from The Rice Foundation. We would like to acknowledge those institutions whose veterinarians and animal management staff took the time to search their records and fill out this survey. The authors would also like to thank research assistant Jane Fouser for her time and energy in every phase of this project.
Literature Cited
1. Ball, S., M. Brown, P. Daszak and R. Pittilo. 1998. Atoxoplasma (Apicomplexa: Eimeriorina: Atoxoplasmatidae) in the greenfinch (Carduelis chloris). J. Parasitol. 84(4):813–817.
2. Collins, M.S., T.B. Smith, R.E. Seibels and I Made Wedana Adi Putra. 1998. Approaches to the reintroduction of the Bali mynah. Zoo Biology. 17:267–284.
3. Daszak, P. and D. Jeggo. Atoxoplasmosis in the Bali mynah in the UK & the taxonomic status of Atoxoplasma. 1999 Proceedings of the Wildlife Disease Conference. Athens, GA. Pp. 94.
4. EpiInfo. 2000. Centers for Disease Control. Atlanta, GA. www.cdc.gov/epiinfo/ (date accessioned 11/15/2000).
5. Greiner, E.C., T.M. Norton, K.S. Latimer and R. Seibels. Atoxoplasmosis - an impediment to the Bali mynah (Leucopsar rothschildi). 1995 Proceedings American Association of Zoo Veterinarians. Pp. 211.
6. Loomis, M.R. and J.F. Wright. Treatment of iron storage disease in a Bali mynah. 1993 Proceedings American Association of Zoo Veterinarians. Pp. 28.
7. Norton, T.M., E.C. Greiner, K.S. Latimer and E. Dierenfeld. Medical aspects of Bali mynahs (Leucopsar rothschildi). 1993 Proceedings American Association of Zoo Veterinarians. Pp. 29–37.
8. Norton, T.M., E.C. Greiner, K.S. Latimer and S.E. Little. 2000. Medical protocols recommended by the U.S. Bali mynah SSP. www.riverbanks.org/aig. (VIN editor: link could not be accessed on 2/25/21)
9. Norton, T.M., R.E. Seibels, E.G. Greiner and K.S. Latimer. 1995. Bali mynah (Leucopsar rothschildi) captive medical management and reintroduction program. Proceedings of the Association of Avian Veterinarians. Pp. 38.
10. Partington, C.J., C.H. Gardiner and D. Fritz. 1989. Atoxoplasma in Bali mynahs. J. Zoo Wildl. Med. 20(3):328–335.
11. Seibels, R., S. Thompson and S. Long. 2000. Complete master plan update: Bali mynah Species Survival Plan. American Zoo and Aquarium Association Population Management Center in Chicago. Pp.1.