Recombinant F1-V fusion protein protects black-footed ferrets (Mustela nigripes) against virulent Yersinia pestis infection.

Abstract

Black-footed ferrets (Mustela nigripes) are highly susceptible to sylvatic plague, caused by the bacterium Yersinia pestis, and this disease has severely hampered efforts to restore ferrets to their historic range. A study was conducted to assess the efficacy of vaccination of black-footed ferrets against plague using a recombinant protein vaccine, designated F1-V, developed by personnel at the U.S. Army Medical Research Institute of Infectious Diseases. Seven postreproductive black-footed ferrets were immunized with the vaccine, followed by two booster immunizations on days 23 and 154; three control black-footed ferrets received a placebo. After the second immunization, antibody titers to both F1 and V antigen were found to be significantly higher in vaccinates than controls. On challenge with 7,800 colony-forming units of virulent plague by s.c. injection, the three control animals died within 3 days, but six of seven vaccinates survived with no ill effects. The seventh vaccinate died on day 8. These results indicate that black-footed ferrets can be immunized against plague induced by the s.c. route, similar to fleabite injection.

Full text

Page 1

142
Journal of Zoo and Wildlife Medicine 35(2): 142–146, 2004
Copyright 2004 by American Association of Zoo Veterinarians
RECOMBINANT F1-V FUSION PROTEIN PROTECTS
BLACK-FOOTED FERRETS (MUSTELA NIGRIPES) AGAINST
VIRULENT YERSINIA PESTIS INFECTION
Tonie E. Rocke, M.S., Ph.D., Jordan Mencher, D.V.M., M.S., Susan R. Smith, B.S.,
Arthur M. Friedlander, M.D., Gerard P. Andrews, Ph.D., and Laurie A. Baeten, D.V.M.
Abstract: Black-footed ferrets (Mustela nigripes) are highly susceptible to sylvatic plague, caused by the bacterium
Yersinia pestis, and this disease has severely hampered efforts to restore ferrets to their historic range. A study was
conducted to assess the efficacy of vaccination of black-footed ferrets against plague using a recombinant protein
vaccine, designated F1-V, developed by personnel at the U.S. Army Medical Research Institute of Infectious Diseases.
Seven postreproductive black-footed ferrets were immunized with the vaccine, followed by two booster immunizations
on days 23 and 154; three control black-footed ferrets received a placebo. After the second immunization, antibody
titers to both F1 and V antigen were found to be significantly higher in vaccinates than controls. On challenge with
7,800 colony-forming units of virulent plague by s.c. injection, the three control animals died within 3 days, but six
of seven vaccinates survived with no ill effects. The seventh vaccinate died on day 8. These results indicate that black-
footed ferrets can be immunized against plague induced by the s.c. route, similar to fleabite injection.
Key words: Black-footed ferret, Mustela nigripes, plague, vaccination, Yersinia pestis.
INTRODUCTION
Sylvatic plague, caused by the bacterium Yer-
sinia pestis, is primarily a disease of wild rodents
that is transmitted by fleas and can afflict humans
as well as other mammals. The plague bacterium
was probably introduced into the United States
through seaports in the early 1900s and quickly
spread into native rodent populations, particularly
in the western states.3The disease was first ob-
served in prairie dogs (Cynomys spp.) in the 1930s
in Arizona and New Mexico6and since then has
had devastating effects on many prairie dog pop-
ulations, often killing 90–100% of individuals in
affected colonies.9Today, sylvatic plague has
spread throughout the range of Gunnison’s prairie
dogs (Cynomys gunnisoni) in Arizona, Utah, New
Mexico, and Colorado. It also persists in white-
tailed prairie dogs (C. leucurus) in Wyoming and
Colorado, in Utah prairie dogs (C. parvidens) in
Utah, and in black-tailed prairie dogs (C. ludovi-
cianus) in Colorado, New Mexico, Texas, Ok-
lahoma, and Montana.2,5
The endangered black-footed ferret (Mustela ni-
From the U.S. Geological Survey, Biological Resources
Division, National Wildlife Health Center, 6006 Schroeder
Road, Madison, Wisconsin 53711, USA (Rocke, Mencher,
Smith, Baeten); and U.S. Army Medical Research Institute
of Infectious Diseases, Bacteriology Division, Fort De-
trick, Frederick, Maryland 21702, USA (Friedlander, An-
drews). Present address (Baeten): Laboratory Animal Re-
sources, Painter Center, Colorado State University, Fort
Collins, Colorado 80525, USA. Correspondence should be
directed to Dr. Rocke.
gripes) is also highly susceptible to plague and may
suffer high mortality rates on infection.10Other car-
nivores may become infected with Y. pestis and se-
roconvert, but with the exception of felines, they
are somewhat resistant to the disease. This includes
domestic ferrets (Mustela putorius furo) that have
survived experimental injections of large numbers
of the bacteria.10Black-footed ferrets depend pri-
marily on prairie dogs for both food and shelter and
thus may be exposed to the bacteria either by con-
sumption of plague-infected prey or by fleabite.
Once thought to be extinct, a captive breeding and
recovery program was established for black-footed
ferrets in 1987 after an outbreak of canine distem-
per nearly decimated the last known wild colony
that was discovered 6 yr earlier. The occurrence of
plague in prairie dog populations and its potentially
devastating effect on black-footed ferret reestab-
lishment is a major impediment to this recovery
program.3
Although attempts have been made, with limited
success, to control plague in prairie dog colonies
by dusting burrows with insecticides after the onset
of an epizootic or through population reduction,
neither of these methods are appropriate for an en-
dangered species such as the black-footed ferret.
Recent studies have shown that multiple doses of a
recombinant vaccine, consisting of two fused
plague antigens, F1 and V (F1-V fusion protein),
protect mice against the bubonic or pneumonic
form of plague.8Our objectives were to determine
whether F1-V antigen administered through s.c. in-
jection could protect black-footed ferrets from vir-
ulent plague challenge.

Page 2

143
ROCKE ET AL.—VACCINATION OF FERRETS AGAINST PLAGUE
MATERIALS AND METHODS
Experimental animals
Black-footed ferrets (five males and five females)
were obtained from the U.S. Fish and Wildlife Ser-
vice’s Black-footed Ferret Recovery Program and
transported to the U.S. Geological Survey National
Wildlife Health Center (NWHC), Madison, Wis-
consin, where they were placed in a Biosafety Lev-
el 3 animal-holding facility. The black-footed fer-
rets came from multiple captive breeding facilities
including the National Zoological Park’s Conser-
vation Research Center, Louisville Zoological Gar-
dens, the Cheyenne Mountain Zoo, and the Nation-
al Black-footed Ferret Conservation Center. All an-
imals were 3–5 yr of age, had been surgically ster-
ilized, and were vaccinated against rabies and
canine distemper.
The animals were housed individually in stain-
less steel rabbit cages (76 ? 51 ? 41 cm). Cage
doors were covered with Plexiglas? acrylic sheets
to prevent escape from the cage. Food and water
were provided ad lib. in stainless steel boxes and
sipper bottles. Nest boxes (31 ? 15 ? 23 cm; Rub-
bermaid Home Products, 1147 Akron Road, Woo-
ster, Ohio 44691, USA) containing shredded paper,
20-cm polyvinyl chloride elbows, 30-cm corrugated
drainpipe, and plastic balls were placed into each
cage. On arrival at NWHC, the animals were
weaned from a raw meat diet (Toronto Carnivore
Mix Diet, Milliken Meat Products, 3347 Kennedy
Road #1, Scarborough, Ontario M1V3F1, Canada)
and acclimated to a pelleted ferret diet (Totally Fer-
ret?, Performance Foods Inc., 3001 Industrial Lane,
Unit 4, Broomfield, Colorado 80020, USA) over a
4- to 6-wk period. Additional food items included
Iams? canned cat food (Iams Company, 7250 Poe
Avenue, Dayton, Ohio 45414, USA; 15 ml, 1–3
times per week) and thawed mice or pinkies (1–4
times per week).
Prophylactic treatment with Albon? (sulfadime-
thoxine; Pfizer Inc., 235 East 42nd Street, New
York, New York 20017, USA; p.o., 55 mg/kg) was
administered at shipment, followed by 22.5 mg/kg
once daily for 10 days after shipment to prevent
coccidiosis. An additional course of treatment with
trimethoprim/sulfa (Tribrisen? Schering-Plough
Corporation, World Headquarters, 2000 Galloping
Hill Road, Kenilworth, New Jersey 07033, USA;
30 mg/kg, p.o. for 14 days) was required during the
acclimation period. For handling and restraint,
black-footed ferrets were chamber induced with is-
oflurane (AErrane?, Baxter, Deerfield, Illinois
60015, USA) administered at 5% with a flow rate
of 2 L oxygen/min. Once induced, the animals were
maintained in a state of anesthesia with 2–3% iso-
flurane. Blood samples were collected from jugular
veins using 23 ga needles with 3-cc syringes.
Vaccine preparation
The F1-V plague vaccine has been described.8
Before administration to black-footed ferrets, the
antigen was diluted in modified Dulbecco medium
(Sigma Chemical Co., P.O. Box 14508, St. Louis,
Missouri 63178, USA) to provide a concentration
of 400 ?g/ml. This solution was mixed 1:1 with an
adjuvant, 0.2% Alhydrogel (United Vaccines, 7819
Airport Road, Madison, Wisconsin 53562, USA),
that was also diluted in modified Dulbecco medi-
um. The resulting vaccine–adjuvant mixture was
rocked gently overnight at 4?C.
Yersinia pestis preparation and isolation
The strain, Y. pestis CO92, a wild-type isolate of
human origin, was provided by the U.S. Army
Medical Research Institute of Infectious Diseases
(USAMRIID, Fort Detrick, Frederick, Maryland
21792, USA). Seventy-five microliters of the iso-
late was spread on blood agar plates (Remel, 12076
Sante Fe Drive, Lenexa, Kansas 66215, USA) and
incubated for 48 hr at 28?C. A heavy culture was
obtained by scraping the growth, inoculating 200
ml heart infusion broth (Difco Laboratories, De-
troit, Michigan 48201, USA) containing 2% xylose,
and incubating for 48 hr at 28?C. A stock suspen-
sion was prepared by adding sterile glycerol (20%
v/v) to the broth culture, mixing thoroughly, and
aliquoting 1 ml volumes. Aliquots were stored fro-
zen at ?80?C.
Concentration, purity, and virulence of our Y.
pestis stocks were determined by numerous plate
counts and mouse tests. In each trial, serial 10-fold
dilutions of a 1-ml aliquot were made in 0.85%
saline. Concentration was determined by plating
100 ?l per dilution onto blood agar plates and in-
cubating for 48 hr at 28?C. Stock concentration was
based on an average of the plate counts trials,
which were consistent and pure. Virulence was de-
termined by s.c. injection of outbred ICR female
mice (7–12 wk old; Harlan Sprague Dawley Inc.,
P.O. Box 29176, Indianapolis, Indiana 46229,
USA); four mice were inoculated with 0.2 ml of
each dilution. The Reed–Muench method was used
to calculate a mouse 50% lethal dose (LD50) of 20
colony-forming units (cfu) for our challenge stock.
The challenge inoculum for the black-footed fer-
rets was prepared by diluting an entire 1-ml aliquot
to the desired concentration in sterile saline. Serial
dilutions of the inoculum were spread on blood
agar plates to confirm the actual cfu. Plates were

Page 3

144
JOURNAL OF ZOO AND WILDLIFE MEDICINE
incubated for 48 hr at 28?C. Mouse tests were per-
formed to confirm virulence of the challenge in-
oculum.
Isolation of Y. pestis from tissue of challenged
animals was accomplished by direct plating on
blood agar. The liver, lung, and a nasal swab from
each animal were cultured for Y. pestis. From some
animals, the spleen, heart, kidney, and skin tissue
at the site of injection were also cultured depending
on gross observations. Tissue impressions of cut
surfaces were made on blood agar plates and
streaked for isolation. Plates were incubated at 28?C
for 48 hr. Subcultures of presumptive Y. pestis were
tested and identified with bacteriophage strips ob-
tained from the Centers for Disease Control and
Prevention (Fort Collins, Colorado 80522, USA).
Subcultures of other bacteria were identified using
the API 20E system (bioMerieux Vitek Inc., 595
Anglum Drive, Hazelwood, Missouri 63042, USA).
Experimental design
This study was reviewed and approved by the
NWHC’s Animal Care and Use Committee and
Biosafety Committee. All personnel handling
plague-infected animals or carcasses were required
to wear powered air-purifying (Hepa-filtered) res-
pirators with full-face shields, rubber aprons and
boots, and double surgical gloves. In addition, per-
sonnel collecting and handling animals and con-
ducting necropsies were required to take prophy-
lactic antibiotics (as prescribed by occupational
health physicians).
Seven black-footed ferrets received 0.5-ml F1-V
vaccine–adjuvant preparation (100 ?g of antigen)
by s.c. injection between the scapulae. Three con-
trol animals received a placebo of 0.5-ml Dulbecco
medium. On day 23, all animals were immunized
again with the same doses of vaccine and placebo,
as described above, and a third immunization was
administered on day 154. Three weeks later (day
176), the animals were challenged with 7,800 cfu
of Y. pestis (at least 10 LD50) administered in 0.2-
ml sterile saline by s.c. injection in the right caudal
thigh. Blood samples were taken from animals be-
fore first vaccination and on days 23, 50, and 169.
Animals were monitored daily for signs of illness,
and day of death was noted; severely debilitated
animals were euthanized. Ferrets surviving to 26
days after challenge were bled and then euthanized
by intracardiac injection of euthanasia solution (Eu-
thasol?; Delmarva Laboratories Inc., 1500 Hugue-
not Road, Suite 106, Midlothian, Virginia 23113,
USA). Dead or euthanized animals were frozen at
?20?C and later necropsied on a selected day to
minimize human exposure to Y. pestis. Selected tis-
sues were collected for bacterial isolation.
Serology
Blood samples were collected in sterile glass se-
rum separator tubes and centrifuged. Serum was
transferred to 2-ml polypropylene tubes and frozen
at ?20?C until assayed. Antibody to F1 was ana-
lyzed at both NWHC and USAMRIID using an en-
zyme-linked immunosorbent assay (ELISA). Se-
lected samples were also analyzed for antibody to
V antigen at USAMRIID by ELISA. Briefly, 96-
well polystyrene plates (Polysorp Nunc-Immuno
plates; Nalge Nunc International, 75 Panorama
Creek Drive, P.O. Box 20365, Rochester, New York
14602, USA) were coated overnight at 4?C with
0.7-?g F1 antigen (USAMRIID) in a 100-?l vol-
ume or 0.1 ?g of V antigen (USAMRIID) in a 50-
?l volume. Serum samples serially diluted from 1:
50 to 1:12,800 in TRIS-buffered saline with Tween
20 were added to wells in 50 ?l volumes and in-
cubated at 37?C for 1 hr. Goat anti-domestic ferret
immunoglobulin G (IgG) conjugated to horseradish
peroxidase (KPL, 2 Cessna Court, Gaithersburg,
Maryland 20879, USA) and diluted 1:80 was added
to wells (50 ?l/well) and incubated at 37?C for 1
hr. Plates were washed four times in TRIS-buffered
saline with Tween 20 after each of these steps. Per-
oxidase (ABTS Peroxidase Substrate?, KPL) was
used as the substrate and 1% sodium dodecyl sul-
fate as the stop solution. Optical absorbance was
measured at 405 nm using an EL800 plate reader
(Bio/Tek Instruments Inc., P.O. Box 998, Winooski,
Vermont 05404, USA).
Statistical analysis
Antibody titers were transformed by calculating
the log10of the reciprocal titer value. Change in
titer was then calculated by subtracting an individ-
ual animal’s transformed preinoculation anti-F1 or
anti-V titer from the transformed titer of each of
that same animal’s subsequent blood samples. Sta-
tistical difference in change of titer between groups
was tested separately at each blood sampling period
using a one-tailed Mann–Whitney test11at P ?
0.05. Difference in survivorship between groups
was tested at P ? 0.05 using the Fisher Exact test,11
and days to death were compared using a one-tailed
Mann–Whitney test at P ? 0.10.
RESULTS
Of the seven black-footed ferrets immunized
with F1-V, six developed an increased level of anti-
F1 IgG antibodies on immunization compared with
prevaccination baseline (Fig. 1); one animal did not

Page 4

145
ROCKE ET AL.—VACCINATION OF FERRETS AGAINST PLAGUE
Figure 1.
noglobulin G antibody titer in response to vaccination.
Solid lines represent responses of vaccinated animals;
dashed lines represent responses of unvaccinated animals.
The vaccinate that died is indicated by the solid circle.
Change in log10reciprocal anti-F1 immu-
Figure 2.
G antibody titer in response to vaccination. Solid lines
represent responses of vaccinated animals; dashed lines
represent responses of unvaccinated animals. The vacci-
nate that died is indicated by the solid circle.
The log10reciprocal anti-V immunoglobulin
develop F1 antibody until after the challenge. None
of the negative controls showed an increase in titer
over baseline at any point. The mean change in
antibody titer was statistically significant between
vaccinates and negative controls on day 50 (P ?
0.025) and day 169 (P ? 0.025) but not on day 23
(P ? 0.10) after vaccination. Antibody to V antigen
was detected in three vaccinates on day 23 and in
all seven on day 50, 27 days after the second boost,
and the mean change in titer between vaccinates
and controls was significantly different both on
days 23 and 50 (P ? 0.05; Fig. 2). Anti-V titers
were not run on day 169.
On challenge with virulent plague by s.c. injec-
tion, the three controls died on day 3 or 4, and Y.
pestis was cultured from their tissues. In contrast,
six of seven vaccinates survived plague challenge
(P ? 0.033), with the seventh animal dying on day
8. A mixed culture of Y. pestis and Escherichia coli
was isolated from the lung, liver, and spleen of this
animal. No Y. pestis or other significant bacteria
were detected in the cultured tissues of the surviv-
ing vaccinates that were later euthanized.
On gross examination, control animals were in
good flesh, but the lungs were mottled with pale
and congested areas. Histopathologic examination
revealed numerous bacilli in all tissues examined.
The absence of an inflammatory response suggest-
ed an overwhelming peracute infection. In the one
vaccinate that died, there was a 1-cm nodular, cyst-
like structure in the liver, a 4 ? 12 mm area of
congestion in the stomach mucosa, and thick, red-
dish purple fluid in the gall bladder. Histopathology
revealed that this animal did not have the extensive
disseminated bacteremia observed in the controls
and that an inflammatory response was evident in
numerous tissues. Pneumonia and cholecystitis
were the most severe lesions, but myositis, spleni-
tis, hepatitis, and gastritis were also evident.
DISCUSSION
Black-footed ferrets vaccinated with three doses
of F1-V antigen were protected against s.c. chal-
lenge with virulent Y. pestis. The one vaccinate that
died on plague challenge survived longer (8 days
after challenge) than corresponding controls (3–4
days after challenge). All the ferrets vaccinated
with F1-V antigen developed a noticeable rise in
anti-V antibody after immunization, and all but one
also developed anti-F1 antibody; this animal sur-
vived infection. In mouse studies, the V antigen has
been shown to be a determinant of virulence in Y.
pestis and is also a protective immunogen,1whereas
F1 antigen (capsular protein) does not appear to be
directly involved in virulence of Y. pestis.7Because
at least one of the surviving vaccinates in our ex-
periment only had antibody to V and not F1, anti-
V antibody may be a better indicator of protection
in ferrets than anti-F1.
Black-footed ferrets are extremely sensitive to
plague infection and in nature could be exposed
either by fleabite or consumption of infected prey.
In a previous study, s.c. injection of four black-
footed ferrets with approximately 800 cfu of Y. pes-
tis killed all the exposed animals (Williams, pers.
comm.). The majority of our immunized ferrets sur-
vived s.c. injection of approximately 10 times that
dose (7,800 cfu). Furthermore, in a subsequent pilot
study, two of five ferrets immunized with F1-V sur-
vived consumption of a plague-infected mouse
(Rocke, unpubl. data), a much more potent chal-
lenge with millions of bacteria. Although these re-
sults are promising, further studies with larger sam-
ple sizes are necessary to determine the maximum
plague challenge dose that immunized animals can
survive through different routes of exposure, to de-
termine the duration of immunity after vaccination,
and to develop an oral route of administration that
can be used in the field. The animals used in this

Page 5

146
JOURNAL OF ZOO AND WILDLIFE MEDICINE
study were postreproductive, had been moved from
other facilities, and were 3–5 yr old. Several ani-
mals developed tumors or other problems during
the course of our study. Higher antibody titers and
an improved response to challenge may be obtained
on vaccination of younger animals (1–3 yr) and an-
imals housed in more natural environments.
Until other methods of plague control are devel-
oped, the F1-V vaccine might be useful in protect-
ing animals in captive-breeding facilities and ani-
mals intended for release programs. Black-footed
ferret kits and dams in captive breeding programs
are fed wild prairie dogs that are captured and
killed for that purpose. However, the loss of nu-
merous captive ferrets at one facility from ingestion
of plague-infected meat demonstrated the hazard of
this practice even with quarantine of the prairie
dogs.4Vaccination of captive black-footed ferrets
against plague could reduce this risk. Ferrets in-
tended for release into the wild could be immunized
with F1-V antigen several times before release and
reimmunized on recapture. This might reduce mor-
tality rates of ferrets during plague outbreaks. How-
ever, control of plague in black-footed ferrets and
the ultimate recovery of the species will require
control of the disease in their primary prey, prairie
dogs.
Acknowledgments: We thank M. Lockhart and P.
Marinari for assistance in obtaining animals for this
study and L. Sileo for conducting necropsies. We
thank J. Bayerl, C. Bolt, J. Dein, R. Johnson, R.
LaRose, P. Nol, and P. Vertz for animal care and
technical assistance and P. Nol and J. Osorio for
critical review of the manuscript. Funding was pro-
vided by the U.S. Fish and Wildlife Service, U.S.
Biological Resources Division, National Fish and
Wildlife Foundation, Turner Foundation, and the
National Wildlife Federation.
LITERATURE CITED
1. Anderson, G. W., Jr., S. E. C. Leary, E. D. William-
son, R. W. Titball, S. L. Welkos, P. L. Worsham, and A.
M. Friedlander. 1996. Recombinant V antigen protects
mice against pneumonic and bubonic plague caused by
F1-capsule-positive and -negative strains of Yersinia pes-
tis. Infect. Immun. 64: 4580–4585.
2. Barnes, A. M. 1982. Surveillance and control of
plague in the United States. In: Edwards, M. S., and U.
McDonnell (eds.). Animal Disease in Relation to Animal
Conservation. Symposia of the Zoological Society of Lon-
don 50. Academic Press, New York. Pp. 237–270.
3. Barnes, A. M. 1993. A review of plague and its rel-
evance to prairie dog populations and the black-footed
ferret. In: Oldemeyer, J. L., D. E. Biggins, B. J. Miller,
and R. Crete (eds.). Proceedings of the Symposium on the
Management of Prairie Dog Complexes for the Reintro-
duction of the Black-Footed Ferret. U.S. Department of
the Interior Fish and Wildlife Service. Washington, D.C.
Biol. Rep. 13, Pp. 28–37.
4. Castle, K. T., D. Biggins, L. G. Carter, M. Chu, K.
Innes, and J. Wimsatt. 2001. Susceptibility of the Siberian
polecat to subcutaneous and oral Yersinia pestis exposure.
J. Wildl. Dis. 37: 746–754.
5. Cully, J. F., and E. S. Williams. 2001. Interspecific
comparisons of sylvatic plague in prairie dogs. J. Mam-
mal. 82: 894–505.
6. Eskey, C. R., and V. H. Haas. 1940. Plague in the
western part of the United States. U.S. Public Health Ser-
vice, Publ. Health Bull. 54: 1–83.
7. Friedlander, A. M., S. L. Welkos, P. L. Worsham, G.
P. Andrews, D. G. Heath, G. W. Anderson, Jr., M. L. M.
Pitt, J. Estep, and K. Davis. 1995. Relationship between
virulence and immunity as revealed in recent studies of
the F1 capsule of Yersinia pestis. Clin. Infect. Dis. 21:
S178–S181.
8. Heath, D. G., G. W. Anderson, Jr., J. M. Mauro, S.
L. Welkos, G. P. Andrews, J. J. Adamocivz, and A. M.
Friedlander. 1998. Protection against experimental bubon-
ic and pneumonic plague by a recombinant capsular F1-
V antigen fusion protein. Vaccine 16: 1131–1137.
9. Rayor, L. S. 1985. Dynamics of a plague outbreak
in Gunnison’s prairie dog. J. Mammal. 66: 194–196.
10. Williams, E. S., K. Mills, D. R. Kwiatkowski, E.
T. Thorne, and A. Boerger-Fields. 1994. Plague in a black-
footed ferret (Mustela nigripes). J. Wildl. Dis. 30: 581–
585.
11. Zar, J. H. 1999. Biostatistical Analysis, 4th ed.
Prentice Hall, Upper Saddle River, New Jersey.
Received for publication 20 February 2003