This article was downloaded by: [New York University] On: 11 April 2015, At: 16:04 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
British Poultry Science Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/cbps20
Effects of strain of chickens and vaccination with turkey herpesvirus on marek's disease and lymphoid Leukosis in breeding stocks a
a
a
J. S. Gavora , E. S. Merritt , A. A. Grunder & R. S. Gowe
a
a
Animal Research Institute, Research Branch , Agriculture Canada, Ottawa, Ontario, KrA OC6, Canada Published online: 08 Nov 2007.
To cite this article: J. S. Gavora , E. S. Merritt , A. A. Grunder & R. S. Gowe (1975) Effects of strain of chickens and vaccination with turkey herpesvirus on marek's disease and lymphoid Leukosis in breeding stocks , British Poultry Science, 16:4, 375-388, DOI: 10.1080/00071667508416201 To link to this article: http://dx.doi.org/10.1080/00071667508416201
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
Br. Poult. Sci., 16: 375-388. 1975
Longman: printed in Great Britain
EFFECTS OF STRAIN OF CHICKENS AND VACCINATION WITH TURKEY HERPESVIRUS ON MAREK'S DISEASE AND LYMPHOID LEUKOSIS IN BREEDING STOCKS 1 J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
Downloaded by [New York University] at 16:04 11 April 2015
Animal Research Institute, Research Branch, Agriculture Canada, Ottawa, Ontario KrA OC6, Canada Received for publication 30th August 1974
1. A total of 3236 females from eight meat-type strains, half of which were vaccinated for Marek's disease (MD), and 11,193 Leghorn females from ten strains, all vaccinated for MD, were adventitiously exposed to MD and lymphoid leukosis (LL) viruses and observed to 392 and 497 d of age, respectively. 2. In the meat-type birds, vaccination reduced total mortality from 43·4% to 27·1% and mortality due to MD from 16·4% to 5·4% but did not affect mortality and LL (2·9% and 3·4%). 3. In the vaccinated Leghorns total mortality was 11 %, including 2·1% from MD and 1·2% from LL. 4. Significant differences between strains of chickens were found in total mortality, as well as in MD and LL mortality. 5. Strain by vaccination interaction was observed in total rearing and adult mortality, as well as in the MD mortality of adult meat-type females. 6. Leghorn strains with higher rate of egg production and meat-type strains with lower growth rate tended to have better viability.
INTRODUCTION
Before successful vaccines were developed, losses from Marek's disease (MD) represented one of the most important economic factors in the poultry industry. Estimates of the annual losses from MD and lymphoid leukosis (LL) combined were $200,000,000 in the USA alone (Han et al., 1969). Although vaccination for MD substantially decreased losses from MD, the two diseases remain an important economic consideration. Information on the relative importance of MD and LL in the " post MD vaccination era " is scarce. In 17 field trials conducted by Purchase et al. (1971, 1972) MD mortality to 18 months of age averaged 2*8% and ig'2% in vaccinated and non-vaccinated flocks respectively. Mortality from LL, recorded in nine trials, 1
Animal Research Institute, Contribution No. 548. 375
Downloaded by [New York University] at 16:04 11 April 2015
OS
TABLE I
t_
w
Description of Leghorn strains
No. of
Origin
Selection
Size of breeding populations in 1971
O >,
jt
Strain
ICI11 dlCS ICS ICQ.
Base population Derived from strain 3
Year
I
(vacc. for MD) 12:4
3
1299
Derived from strain 4
"969
3
1217
Narrow genetic base population (Ottawa) Seven Canadian R.O.P. unrelated stocks
4
1341
5 6
946
7
996
8
Females
28
224
28
224
V
28
224
s
28
224
None None
80 80
240 240
None
None
80
240
High hen-housed egg production High egg production rate High egg production rate
Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability
28
224
28
224
20
80
Secondary traits Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size
1968
None None
i960 1969
•1950
1951
Ci 0 SSCCL
'95°
1400
Same as strain 3 Imported from USDA (Purdue), Cornell Control Four commercial strains crossed ' Derived from strain 7
9
1500
Derived from strain 7
1969
IO
320
Cross of strains 3 and 4 in 1956, selected at Agr. Canada Res. Stn., Agassiz, B.C., imported to Ottawa
1971
99°
Males
Primary trait High egg production rate High egg production rate High hen-housed egg production High hen-housed egg production
1971
w
1 • 5! W
O
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
377
averaged 3*3% and 4*6% in vaccinated and non-vaccinated birds respectively and exceeded io% on two of the farms under observation. Although chickens from eight different commercial breeders were used, the effects of genotype were not evaluated. Other studies dealing with effectiveness of vaccination for MD (reviewed by Purchase, 1973) do not present data on the proportion of economic loss attributable to MD and LL separately. The object of this study was to obtain information on the economic importance of mortality due to MD and LL in chickens vaccinated and non-vaccinated for MD and to examine differences between strains and the interaction of strain and vaccination in mortality. MATERIALS AND METHODS
Downloaded by [New York University] at 16:04 11 April 2015
Leghorn strains
A description of the strains of chickens used is given in Table 1. Strains 3, 4, 5 and 7 have been described by Gowe et al. (1973). The history and the method of maintaining the Cornell Control population (strain 6) was described by King et al. (1959)Three unselected control strains and seven strains that had been selected for high egg production and related traits were maintained from hatching to 497 d of age on an experimental farm where they were adventitiously exposed to MD and LL viruses. A total of 11,193 pedigreed female chicks were hatched in 1971 over 2 consecutive days and all were vaccinated on the day of hatch with a commercial preparation of cell-associated herpesvirus of turkeys. They were reared to 20 weeks in one house in group cages. Gages were assigned randomly to each strain, and progeny of each sire were randomised to each cage assigned to that strain. During the 20th week of age all pullets were moved into individual 20-cm modified stairstep cages assigned to them at random in two adjacent poultry houses. All mash diets were fed ad libitum. The populations were reared under a constant 6-h day (dim red lights) and starting at 17 weeks, daylength was increased by 30 min per week until the birds received a 16-h day which was maintained to 497 d of age (Gowe et al., 1973). Besides vaccination for MD the birds were also vaccinated for both Newcastle disease and infectious bronchitis at 7 and 15 weeks and for avian encephalomyelitis at 14 weeks. Meat-type strains
A description of the strains is given in Table 2. All meat-type strains used were derived from strain K (Merritt and Gowe, 1962). . Two unselected control strains and six body weight selected meat-type strains (a total of 3236 female chicks) were studied from day of hatch to 392 d of age. In 1971, two pedigree hatches were produced 3 weeks apart in all strains and then were maintained on the same farm (but not in the same buildings) as the Leghorns. In both hatches female chicks within each dam family were divided at random into two groups. One of the groups was vaccinated on the day of hatch with commercial cell-associated turkey herpesvirus vaccine. All chicks were reared intermingled in floor pens, each hatch occupying one wing of the same building. At 20 weeks of age all pullets were 16/4—D
378
J. S. GAVORA, E. S. MERR1TT, A. A. GRUNDER AND R. S. GOWE
transferred into single 25-cm stair-step cages assigned to them at random in one poultry house. Adventitious exposure of the birds to the MD and LL viruses was expected to be similar to that in the Leghorns. All mash diets were fed ad libitum. A light regime of 14 h/d was used for the entire test. All females were vaccinated for both Newcastle disease and infectious bronchitis at 2 and 16 weeks and for avian encephalomyelitis at 13 weeks. TABLE 2
Description of meat-type strains
Size of breeding populations in
No. of females tested
Downloaded by [New York University] at 16:04 11 April 2015
i
Origin
I;O*7I1
y/
Strain A
forMD
N % vacc.
"9
121
Base population Strain K
Year 1958
B
168
167
Strain K.
1958
C
197
183
Strain K.
1958
D
180
164
Strain K
1958
E
187
170
Strain K
1958
H
164
182
Strain F
1965
F
341
317
1963
K
290
286
Crossing of strains A, B, andC Four commerical broiler strains crossed
Selection High broiler body weight High broiler body weight High broiler body weight High broiler, low adult body weight High adult body weight High broiler body weight None
1957
None
Male Female 32
79
38
86
49
97
24
64
21
64
29
63
96
218
78
243
Post-mortem examination
The procedure for post-mortem examination was based on the guidelines suggested by Siccardi and Burmester (1970) and it was also consistent with recommendations of Purchase and Sharma (1973). AH birds dying during the test were examined for gross lesions. Briefly, birds with LL were always over 19 weeks of age and had tumours in the bursa of Fabricius. Some also had lesions in the viscera and muscles. Birds considered to have MD were of any age and had visceral, muscle and/or neural lesions but did not have tumours in the bursa of Fabricius. Because diagnoses were based on gross examination, it is possible that some bursal tumours due to LL were not noticed and that some of the bursal lesions were associated with MD (Payne and Rennie, 1970). Percentage mortality in the first five 28-d periods was based on the number of birds hatched. At the end of the fifth period, when the pullets were transferred into the laying houses, their number was slightly reduced but in the reduction no attention was paid to the health status of the birds. From the sixth 28-d period onward the mortality percentages were based on the number of birds housed in the laying house.
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
379
Statistical analyses
The all-or-none data on mortality in the meat strains tested were arranged into four-dimensional contingency tables of counts and were analysed by the method described by Fienberg (1970), using an iterative procedure to fit models by maximum likelihood. One of the variables was considered as a " response " factor always expressed as a dichotomy (e.g. for total mortality the response categories were " dead " or " alive " ) , while the remaining variables (hatch, vaccination and strain) were regarded as " design " variables since the marginal totals for these categories were predetermined before the experiments were initiated. Using the same notation as Fienberg (1970), the sequence of models tested started from model log em
= [i] + [ 4 + [B]j + [C]k +[D]i + [CD]kl + [BD]n + [BC]}k + [AD]U + [AC]ik + [AB]{j + [BCD]m + [ACD]m + [ABD]m + [ABC]m,
(1)
Downloaded by [New York University] at 16:04 11 April 2015
where: eijki = expected count in the ith hatch, jth category of vaccination, kth. strain and /th category of response, [1] = the grand mean of the logarithms of expected counts, [A]i, [B]}, [C]fc and [Z)]j = parameters representing deviations from the grand mean due to the hatch, vaccination, strain and response categories respectively, [CD]ici, [BD]}i,..., [ABC]i]h = two-factor and three-factor effects or interactions.
Subsequent models in the sequence were formed by omitting terms from equation (1) but retaining all parameters involving only " design " variables. Thus the simplest model that could be considered would be: log em
= [i] + [Ay + [B], + [C]k + [D]t + [AB]fi + [AC]ik + [BC]Jk + [ABC]m.
No term may be omitted from the model unless all higher-order interactions involving that term have also been omitted. The adequacy of a particular model was examined using a log-likelihood ratio statistic (LLR). By differencing the LLR values for a pair of successive models, one model generated from the other (the " reference model ") by dropping the parameters for a particular interaction, an LLR value corresponding to that source of variation was obtained. Provided the reference model had given an adequate fit to the data this LLR value was compared with x2 tables to obtain a test of significance. In this way, the LLR for equation (1) may be partitioned into components as in an analysis of variance. When a stage was reached at which no non-significant term could be dropped from the model and at which the model gave an adequate fit, that model was called the " final model ". It should be noted that the sequence of models used is not unique and, because of the sequential procedure, the test of significance for any component, or source of variation, after the first should be interpreted with the usual caution. The same general procedure could be used to assess the effect of strain on mortality in the vaccinated Leghorns and in the vaccinated and non-vaccinated meat-type birds separately. Instead, the analogous procedure of analysis of variance for categorical data in 2-way tables (Catanova) described by Light and Margolin (1971) was used. A feature of the Catanova analysis is the R 2 statistic, which
CO CO
Downloaded by [New York University] at 16:04 11 April 2015
o
TABLE 3
Dependence of mortality in meat-type females on hatch, vaccination and strain of chickens. Results of filing models to the data by maximum likelihood
Adult mortality
Rearing mortality Marek's disease
Total Source variation Hatch (H) Vaccination (V) Strain (S) HV HS VS HVS Final models: Sources retained LLR for goodness of fit Degrees of freedom 1
Degrees of freedom i i
7
i 7 7 7
LLR1 2-84 I4-753 3
5O-39 2-38
5-o6 15-78*
V, S, VS 23-59 48
SEQ? 4
LLR o-68 18-60** 19-73** o-oi 5-38 4-41 8-40 V,S 18-88 4'
SEQ. 4
LLR 0-42 47-51* 90-36* 0-26 12-61
SEQ. 4
18-45* 14-07 V, S, VS 27-36 48
Log-likelihood statistic corresponding to the dependence of mortality on the factors or interactions. Sequence of dropping the sources of variation from the complete model. 3 Relative to model from which the significant VS interaction was excluded. No test of significance made. * LLR statistic exceeds %" value for P = 0-05. ** LLR statistic exceeds x2 value for P = o-oi. 2
Lymphoid leukosis
Marek's disease
Total
LLR 1-36 63-34s 20763 o-8o
6-oo 15-69* 10-63 V, S, VS 18-79 48
SEQ,
LLR
4
0-02
o-54 62-66** 0-04 11-47 6-13
8-34 S 26-54 40
SEQ 4 6
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
381
measures the proportion of the total variation (in the sense defined by Light and Margolin, 1971) which is accounted for by knowledge of the strain. For large
Downloaded by [New York University] at 16:04 11 April 2015
NON-VACCINATED
.1
8
9
10
11
12
13
MORTALITY:
•
NONSPECIFIC
5J
14
LYMPHOID LEUKOSIS
WITHOUT NEURAL INVOLVEMENT
VACCINATED FOR MD
WITH NEURAL INVOLVEMENT
IU
2 cc o
II 2
3
4
5
6 7 8 9 28-DAY PERIODS
10
11
12
13
14
FIG. 1.—Mortality to 392 d in vaccinated and non-vaccinated meat-type birds, strains and hatches combined. H = housed in laying house.
samples, the log likelihood and Catanova procedures are approximately equivalent and the Catanova has been applied here to illustrate its use.
382
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
RESULTS
Meat-type females
Results of fitting models to the mortality data in meat-type females are shown in Table 3. The final models for total rearing and total adult mortality both contained parameters indicating influence of vaccination and strain of chickens on mortality, as well as interaction of vaccination'and strain with mortality. Hatch and interactions of hatch with vaccination or strain were shown to be unimportant MORTALITY:
•
NONSPECIFIC LYMPHOID LEUKOSIS
MAREK'S DISEASE
Downloaded by [New York University] at 16:04 11 April 2015
4-
WITHOUT NEURAL INVOLVEMENT WITH NEURAL INVOLVEMENT
LJ
8 3 UJ
2-
o
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
28-DAY PERIODS FIG. 2.—Mortality to 497 d in vaccinated Leghorns, all strains combined. H = housing in laying house.
not only in total mortality but also in MD and LL mortality (Table 3). Therefore, mortality from both hatches was combined in Figs. 1 and 3 and hatch was disregarded in the Catanova analyses (Table 4). Vaccination for MD reduced the total mortality in all meat strains combined from 11 % to 7% in the rearing period and from 32*4% to 20-1 % in the adult period. The importance of the effect of genotype was illustrated by the fact that during the rearing period total mortality in vaccinated birds varied amongst strains from 3*1% to 16-7% and in non-vaccinated birds from 4*9% to 18-7% (Fig. 3). In the adult period, when mortality was generally higher, strain differences in total mortality were even more pronounced. The lowest total mortality of adult birds (6*i% and 13*5% in vaccinated and non-vaccinated birds respectively) was observed in the unselected control strain K with the lowest average 56-d body weight. On the other hand, strain B with the third highest average 56-d body weight, had the highest mortality among both vaccinated (38-6%) and non-vaccinated (41*3%) birds (Fig. 3). The results of the analyses of variance for categorical data (Table 4) confirm the significance of the effects of strain on total mortality. However, the proportion
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
383
of variation in total mortality explained by strain was generally low and ranged from 1-5% in the non-vaccinated birds in the rearing period to 6-2% in vaccinated birds during the adult period (Table 4). MORTALITY:
•
NONSPECIFIC
40-
ED
LYMPHOID LEUKOSIS
0-140 DAYS OF AGE
£ 30
MAREK'S DISEASE
UJ
o cc 111
WITHOUT NEURAL INVOLVEMENT
LUI
WITH NEURAL INVOLVEMENT
EB
o.
Downloaded by [New York University] at 16:04 11 April 2015
> 20-
NV
Ioc
o
NV
10-
NV •
V
INV V NV
A (1347)
H (1281)
B (1195)
C (1165)
D (1111)
r
E (1098)
V V
f—inor F (1098)
K (873)
STRAIN (AVERAGE 56-DAY BODY WEIGHT IN GRAMS) 143-392 DAYS OF AGE
40-^
V NV
30' u cc
HI
t
£
20
cc
o 5
IOH
A (1347)
H (1281)
B (1195)
C
D
smi
E F K (1165) (1111) (1098) (1098) STRAIN (AVERAGE 56-DAY BODY WEIGHT IN GRAMS)
(873) ,
FIG. 3.—Mortality in meat-type strains. Strains ranked left to right on average 56-d body weight. V = vaccinated for MD, NV = non-vaccinated.
The fact that vaccination reduced the total mortality to different extent in the different strains is clearly seen especially in the adult period (Fig. 3). For example
384
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
the two strains (H and B) with the second and third highest average 56-d body weight showed little or no reduction in mortality with vaccination. Yet strain A, with the highest 56-d body weight, gave a reduction in total mortality from 2 5 7 % to 12-4%, a reduction of the same order as all the other strains. Differences among strains were greater for vaccinated than for non-vaccinated birds (Fig. 3) and this is reflected in the R 2 values shown in Table 4. Monthly mortality of all meat strains combined shown in Fig. 1 indicates that losses increased from the first to the seventh or eighth 28-day period followed by a slight decline. This trend is observed in both the vaccinated and non-vaccinated birds despite the generally lower mortality of the vaccinates. TABLE 4
Downloaded by [New York University] at 16:04 11 April 2015
Proportion of variation in mortality explained by strain 0/chickens (iJ2) and Catanova statistic (C) for differences between strains Meat-type females (df = 7) Vaccinated for MD
Non-vaccinated
Vaccinated Leghorn females (df = 9)
C
R 2 (%)
C
R 2 (%)
G
R 2 (%)
Rearing mortality: Total Marek's disease
47-8* 2-9
2-9 02
24-4** 20-0**
1-5 1-3
29-8** o
0-3 o
Adult mortality: Total Marek's disease Lymphoid leukosis
8I-I** 25-1** 47-8**
6-2 1-9 3-7
31-3** 20*3** 30-0**
2-6 1-3 2-5
23-1** 56-9** 54'3**
o-6 0-7 °'7
* P g 0-05; ** P g o-oi.
The analysis of data on rearing mortality due to MD indicated the dependence of mortality on vaccination and strain only, while the final model for MD mortality in the adult period also contained a parameter for vaccination by strain (by mortality) interaction (Table 3). Overall mortality attributed to MD was reduced by vaccination from 2*4% to o-6% in the rearing period and from 14-0% to 4-8% in the adult period. The Catanova analyses (Table 4) indicated that the significance of strain effects on rearing mortality attributed to MD resulted from strain differences among non-vaccinates. The strain effects on adult MD mortality were highly significant in both vaccinated and non-vaccinated birds (Table 4). While rearing MD mortality in non-vaccinated birds ranged from 5*1% for strain F to o*6% for strains B, G and D (Fig. 3), it was no greater than i-2% for any strain in vaccinated birds. In non-vaccinated adult birds MD mortality ranged from 20-1% for strain C to 9% for strain A (Fig. 3). Although all strains showed some decrease in MD mortality with vaccination, the reduction was 15% for strain C but less than 2% for strain H. In the non-vaccinated meat-type females MD mortality increased from the 3rd to the 7th month (Fig. 1). The first deaths due to MD occurred in nonvaccinated birds at 69 and 55 d of age in the first and second hatches respectively. In the corresponding vaccinated flock MD mortality was delayed respectively to 101 and 105 d in the first and second hatches and an increase was observed between the 4th and 8th month. In both the vaccinated and non-vaccinated flocks some MD mortality was observed to the end of the experiment.
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
385
Mortality from LL was found only in the adult period. It was influenced significantly by strain of chickens (Table 3) but there was no difference between MD vaccinated (3-4%) or non-vaccinated (2-9%) hens. The strain effect was highly significant in both vaccinated and non-vaccinated birds when analysed separately but the proportion of variation in LL mortality explained by strain was low (Table 4). The time pattern of mortality due to LL in meat-type females was not changed by vaccination (Fig. 1). The first LL mortality was observed at 141 d in the first hatch and at 135 d in the second hatch. MORTALITY: 40-1
LYMPHOID LEUKOSIS
0-140 DAYS OF AGE
Downloaded by [New York University] at 16:04 11 April 2015
30-
MAREK'S DISEASE
u cc U
a. >•
a
NONSPECIFIC
WITHOUT NEURAL INVOLVEMENT WITH NEURAL INVOLVEMENT
GD
•
20-
cc
o
10-
0J
•
1—1 t=3
1 (75)
4 9 8 (74) (73) (73)
c=t_
n
B3
«a
P
1 1
3 2 7 6 10 (73) (72) (71) (70) (69) STRAIN (AVERAGE HEN-DAY PERCENT PRODUCTION FROM FIRST EGG TO 497 DAYS)
5 (66)
40-, 141-497 DAYS OF AGE
30-
20cc
o s 10
1 (75)
3 2 7 6 10 (73) (72) (71) (70) (69) STRAIN (AVERAGE HEN-DAY PERCENT PRODUCTION FROM FIRST EGG TO 497 DAYS)
4 (74)
9 (73)
8 (73)
5 (66)
FIG. 4.—Mortality in vaccinated Leghorn strains. Strains ranked left to right on average hen-day percentage egg production from the first egg to 497 d.
386
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
Downloaded by [New York University] at 16:04 11 April 2015
Leghorn females
The mortality in the vaccinated Leghorn females (Fig. 4) was generally low in all strains. The total mortality to 497 d was 11 % including 2-1 % mortality from MD and 1-2% mortality from LL. Effects of strain were highly significant in all types of mortality examined except for MD mortality in the rearing period (Table 4). However, the R 2 values for the effect of strain, shown in the same table, are even lower than those calculated for the meat-type birds. In the adult period, the unselected control strains 5, 6 and 7 tended to have greater total and MD mortality than all the high egg production selected strains except strain 10 (Fig. 4). Also, control strain 5, which had the lowest egg production rate, had the highest MD mortality of all the strains tested. The monthly mortality pattern in the Leghorns (Fig. 2) was quite similar to that observed in vaccinated meat-type females. The first death from MD occurred in the Leghorns at 77 d of age while the first mortality from LL was observed at 162 d of age. DISCUSSION
The results of this study were generally consistent with previous findings. The repeatedly demonstrated reduction of MD mortality by vaccination (Purchase, 1973) was confirmed by the data on meat-type females. Vaccination delayed the first deaths of meat-type females from MD by 34 and 50 d in the first and second hatch respectively. This seems to support the observation by Spencer and Robertson (1974) that vaccination blocked infection with virulent MD virus for approximately 5 weeks after contact exposure. The importance of genetic resistance for successful vaccination (Spencer etal., 1972, 1974) was clearly indicated in this study. However, contrary to another report (Spencer et al., 1975), vaccination did not seem to reduce the strain variation in MD mortality (Table 4). The proportion of variation in mortality explained by strain was low although the effects of strain on mortality were, in most instances, -highly significant. This illustrates the fact noted by Light and Margolin (1971) that, in large sets of data, weakly related variables can nevertheless exhibit statistically significant dependencies. The data in Fig. 4 suggest that the viability of the Leghorn strains selected for both egg production and viability was higher than in the unselected control strains. These results support previous reports (Von Krosigk et al., 1972; Gavora et al., 1974) that a positive genetic correlation exists between resistance to MD and egg production. Examination of mortality and growth rate of the meat strains in Fig. 3 suggests a negative correlation. Of the two unselected control strains (Table 2) only strain K remained unselected since 1958. The other control strain F had been synthesised in 1963 from strains selected for rapid growth since 1958. The 56-d body weight, as well as mortality of the control strain F were similar to those of the growth-selected strains. On the other hand, the control strain K with a substantially lower 56-d body weight than the selected strains showed the lowest mortality of the vaccinated birds in the rearing period and ranked third best among non-vaccinates (Fig. 3). In the adult period strain K had the lowest total mortality. In MD mortality strain K ranked as the second lowest among vaccinates and as the lowest among non-vaccinates. This supports reports of Friars et al. (1972), Han and Smyth
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
387
Downloaded by [New York University] at 16:04 11 April 2015
(1972) and Gavora et al. (1974) indicating a negative genetic correlation between growth rate and resistance to Marek's disease. However, strain A was an exception since it had the highest 56-d body weight and was relatively resistant to MD (Fig. 3). Thus, it would appear possible to produce rapidly growing strains of chickens with good general viability and resistance to MD and LL. Of special interest are the results of this study on the relative importance of MD and LL in vaccinated meat-type breeders and in vaccinated Leghorns. In chickens vaccinated for MD the total lifetime mortality was 11% in Leghorns and 27% in meat-type females. Approximately one-fifth of this mortality was caused by MD and one-tenth by LL. Hence, losses from MD and LL combined amounted to onethird of total lifetime mortality in vaccinated Leghorn and meat-type hens even with the use of an effective MD vaccine. The continuation of research work aimed at further reduction of these losses would seem desirable. ACKNOWLEDGEMENTS
We gratefully acknowledge the technical assistance of Mr P. Apedaile who carried out the post-mortem examinations. We also wish to thank Mr J . Dickie for his valuable technical assistance and Miss L. Francis, Statistical Research Service, Agriculture Canada, for carrying out the maximum likelihood analyses. The suggestions of Mrs P. Morse from the Statistical Research Service and of Dr J . L. Spencer from the Animal Diseases Research Institute, Hull in Ottawa in preparation of the manuscript are gratefully acknowledged. REFERENCES
FIENBERG, S. E. (1970). The analysis of multidimensional contingency tables. Ecology, 5 1 : 419-432. FRIARS, G. W., CHAMBERS, J . R., KENNEDY, A., AND SMITH, A. D. (1972). Selection for resistance to
Marek's disease in conjunction with other economic traits in chickens. Avian Dis., 16: 2-10.
GAVORA, J . S., GRUNDER, A. A., SPENCER, J . L., GOWE, R. S., ROBERTSON, A. AND SPECKMANN,
G. W. (1974). An assessment of effects of vaccination on genetic resistance to Marek's disease. Poult. Sci., 53: 889-897.
GOWE, R. S., LENTZ, W. E. AND STRAIN, J . H. (1973). Long-term selection for egg production in
several strains of White Leghorns: Performance of selected and control strains including genetic parameters of two control strains. 4th Europ. Poult. Conf., London, 225-245. HAN, F. S. AND SMYTH, J . R. (1972). The influence of growth rate on the development of Marek's disease in chickens. Poult. Sci., 5 1 : 975-985. HAN, F. S., SMYTH, J . R., SEVOIAN, M. AND DICKINSON, F. N. (1969). Genetic resistance to leukosis
caused by the J M virus in the fowl. Poult. Sci., 48: 76-87. KING, S. C., CARSON, J. R. AND DOOLITTLE, D. P. (1959). The Connecticut and Cornell random bred populations of chickens. Wld's Poult. Sci. J., 15: 139-159. LIGHT, R. J . AND MARGOLIN, B. M. (1971). An analysis of variance for categorical data. J. Am. statist. Ass., 66: 534-544. MERRITT, E. S. AND GOWE, R. S. (1962). Development and genetic properties of a control strain of meat-type fowl. XIth Wld's Poult. Congr., Mexico, 66-70. PAYNE, L. N. AND RENNIE, M. (1970). Lack of effect of bursectomy on Marek's disease. J. natn. Cancer. Inst., 4 5 : 387-391. PURCHASE, H. G. (1973). Control of Marek's disease by vaccination. Wld's Poult. Sci. J., 29: 238-250. PURCHASE, H. G., OKAZAKI, W. AND BURMESTER, B. R. (1971). Field trials with the herpes virus of turkeys (HVT) strain FC126 as a vaccine against Marek's disease. Poult. Sci., 50: 775-783.
388
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
PURCHASE, H . G., OKAZAKI, W. AND BURMESTER, B. R. (1972). Long-term field trials with the herpesvirus of turkeys vaccine against Marek's disease. Avian Dis., 16: 57-71. PURCHASE, H . G. AND SHARMA, J . M. (1973). The differential diagnosis of lymphoid leukosis and Marek's disease. Am. Ass. Avian Pathologists, Slide Study Set 3. SICCARDI, F. J . AND BURMESTER, B. R. (1970). The differential diagnosis of lymphoid leukosis and Marek's disease. Tech. Bull. 1412, Washington, Agricultural Research Service, United States Department of Agriculture. SPENCER, J . L., GAVORA, J . S., GRUNDER, A. A., ROBERTSON, A. AND SPECKMANN, G. W. (1974).
Immunization against Marek's disease: Influence of strain of chickens, maternal antibody and type of vaccine. Avian Dis., 18: 33-44. SPENCER, J . L., GRUNDER, A. A., ROBERTSON, A. AND SPECKMANN, G. W. (1972).
Attenuated
Marek's disease herpesvirus: protection conferred to strains of chickens varying in genetic resistance. Avian Dis., 16: 94-107. SPENCER, J . L., AND ROBERTSON, A. (1975). Influence of vaccination with avirulent herpesviruses on subsequent infection of chickens with virulent Marek's disease herpesvirus. Am. J. vet. Res. (in press).
Downloaded by [New York University] at 16:04 11 April 2015
VON KROSIGK, C. M., MCCLARY, C. F., VIELITZ, E. AND ZANDER, D. V. (1972).
Selection for
resistance to Marek's disease and its expected effects on other important traits in White Leghorn crosses. Avian Dis., 16: 11-19.
British Poultry Science Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/cbps20
Effects of strain of chickens and vaccination with turkey herpesvirus on marek's disease and lymphoid Leukosis in breeding stocks a
a
a
J. S. Gavora , E. S. Merritt , A. A. Grunder & R. S. Gowe
a
a
Animal Research Institute, Research Branch , Agriculture Canada, Ottawa, Ontario, KrA OC6, Canada Published online: 08 Nov 2007.
To cite this article: J. S. Gavora , E. S. Merritt , A. A. Grunder & R. S. Gowe (1975) Effects of strain of chickens and vaccination with turkey herpesvirus on marek's disease and lymphoid Leukosis in breeding stocks , British Poultry Science, 16:4, 375-388, DOI: 10.1080/00071667508416201 To link to this article: http://dx.doi.org/10.1080/00071667508416201
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
Br. Poult. Sci., 16: 375-388. 1975
Longman: printed in Great Britain
EFFECTS OF STRAIN OF CHICKENS AND VACCINATION WITH TURKEY HERPESVIRUS ON MAREK'S DISEASE AND LYMPHOID LEUKOSIS IN BREEDING STOCKS 1 J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
Downloaded by [New York University] at 16:04 11 April 2015
Animal Research Institute, Research Branch, Agriculture Canada, Ottawa, Ontario KrA OC6, Canada Received for publication 30th August 1974
1. A total of 3236 females from eight meat-type strains, half of which were vaccinated for Marek's disease (MD), and 11,193 Leghorn females from ten strains, all vaccinated for MD, were adventitiously exposed to MD and lymphoid leukosis (LL) viruses and observed to 392 and 497 d of age, respectively. 2. In the meat-type birds, vaccination reduced total mortality from 43·4% to 27·1% and mortality due to MD from 16·4% to 5·4% but did not affect mortality and LL (2·9% and 3·4%). 3. In the vaccinated Leghorns total mortality was 11 %, including 2·1% from MD and 1·2% from LL. 4. Significant differences between strains of chickens were found in total mortality, as well as in MD and LL mortality. 5. Strain by vaccination interaction was observed in total rearing and adult mortality, as well as in the MD mortality of adult meat-type females. 6. Leghorn strains with higher rate of egg production and meat-type strains with lower growth rate tended to have better viability.
INTRODUCTION
Before successful vaccines were developed, losses from Marek's disease (MD) represented one of the most important economic factors in the poultry industry. Estimates of the annual losses from MD and lymphoid leukosis (LL) combined were $200,000,000 in the USA alone (Han et al., 1969). Although vaccination for MD substantially decreased losses from MD, the two diseases remain an important economic consideration. Information on the relative importance of MD and LL in the " post MD vaccination era " is scarce. In 17 field trials conducted by Purchase et al. (1971, 1972) MD mortality to 18 months of age averaged 2*8% and ig'2% in vaccinated and non-vaccinated flocks respectively. Mortality from LL, recorded in nine trials, 1
Animal Research Institute, Contribution No. 548. 375
Downloaded by [New York University] at 16:04 11 April 2015
OS
TABLE I
t_
w
Description of Leghorn strains
No. of
Origin
Selection
Size of breeding populations in 1971
O >,
jt
Strain
ICI11 dlCS ICS ICQ.
Base population Derived from strain 3
Year
I
(vacc. for MD) 12:4
3
1299
Derived from strain 4
"969
3
1217
Narrow genetic base population (Ottawa) Seven Canadian R.O.P. unrelated stocks
4
1341
5 6
946
7
996
8
Females
28
224
28
224
V
28
224
s
28
224
None None
80 80
240 240
None
None
80
240
High hen-housed egg production High egg production rate High egg production rate
Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability
28
224
28
224
20
80
Secondary traits Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size Fertility, hatchability, viability, egg size
1968
None None
i960 1969
•1950
1951
Ci 0 SSCCL
'95°
1400
Same as strain 3 Imported from USDA (Purdue), Cornell Control Four commercial strains crossed ' Derived from strain 7
9
1500
Derived from strain 7
1969
IO
320
Cross of strains 3 and 4 in 1956, selected at Agr. Canada Res. Stn., Agassiz, B.C., imported to Ottawa
1971
99°
Males
Primary trait High egg production rate High egg production rate High hen-housed egg production High hen-housed egg production
1971
w
1 • 5! W
O
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
377
averaged 3*3% and 4*6% in vaccinated and non-vaccinated birds respectively and exceeded io% on two of the farms under observation. Although chickens from eight different commercial breeders were used, the effects of genotype were not evaluated. Other studies dealing with effectiveness of vaccination for MD (reviewed by Purchase, 1973) do not present data on the proportion of economic loss attributable to MD and LL separately. The object of this study was to obtain information on the economic importance of mortality due to MD and LL in chickens vaccinated and non-vaccinated for MD and to examine differences between strains and the interaction of strain and vaccination in mortality. MATERIALS AND METHODS
Downloaded by [New York University] at 16:04 11 April 2015
Leghorn strains
A description of the strains of chickens used is given in Table 1. Strains 3, 4, 5 and 7 have been described by Gowe et al. (1973). The history and the method of maintaining the Cornell Control population (strain 6) was described by King et al. (1959)Three unselected control strains and seven strains that had been selected for high egg production and related traits were maintained from hatching to 497 d of age on an experimental farm where they were adventitiously exposed to MD and LL viruses. A total of 11,193 pedigreed female chicks were hatched in 1971 over 2 consecutive days and all were vaccinated on the day of hatch with a commercial preparation of cell-associated herpesvirus of turkeys. They were reared to 20 weeks in one house in group cages. Gages were assigned randomly to each strain, and progeny of each sire were randomised to each cage assigned to that strain. During the 20th week of age all pullets were moved into individual 20-cm modified stairstep cages assigned to them at random in two adjacent poultry houses. All mash diets were fed ad libitum. The populations were reared under a constant 6-h day (dim red lights) and starting at 17 weeks, daylength was increased by 30 min per week until the birds received a 16-h day which was maintained to 497 d of age (Gowe et al., 1973). Besides vaccination for MD the birds were also vaccinated for both Newcastle disease and infectious bronchitis at 7 and 15 weeks and for avian encephalomyelitis at 14 weeks. Meat-type strains
A description of the strains is given in Table 2. All meat-type strains used were derived from strain K (Merritt and Gowe, 1962). . Two unselected control strains and six body weight selected meat-type strains (a total of 3236 female chicks) were studied from day of hatch to 392 d of age. In 1971, two pedigree hatches were produced 3 weeks apart in all strains and then were maintained on the same farm (but not in the same buildings) as the Leghorns. In both hatches female chicks within each dam family were divided at random into two groups. One of the groups was vaccinated on the day of hatch with commercial cell-associated turkey herpesvirus vaccine. All chicks were reared intermingled in floor pens, each hatch occupying one wing of the same building. At 20 weeks of age all pullets were 16/4—D
378
J. S. GAVORA, E. S. MERR1TT, A. A. GRUNDER AND R. S. GOWE
transferred into single 25-cm stair-step cages assigned to them at random in one poultry house. Adventitious exposure of the birds to the MD and LL viruses was expected to be similar to that in the Leghorns. All mash diets were fed ad libitum. A light regime of 14 h/d was used for the entire test. All females were vaccinated for both Newcastle disease and infectious bronchitis at 2 and 16 weeks and for avian encephalomyelitis at 13 weeks. TABLE 2
Description of meat-type strains
Size of breeding populations in
No. of females tested
Downloaded by [New York University] at 16:04 11 April 2015
i
Origin
I;O*7I1
y/
Strain A
forMD
N % vacc.
"9
121
Base population Strain K
Year 1958
B
168
167
Strain K.
1958
C
197
183
Strain K.
1958
D
180
164
Strain K
1958
E
187
170
Strain K
1958
H
164
182
Strain F
1965
F
341
317
1963
K
290
286
Crossing of strains A, B, andC Four commerical broiler strains crossed
Selection High broiler body weight High broiler body weight High broiler body weight High broiler, low adult body weight High adult body weight High broiler body weight None
1957
None
Male Female 32
79
38
86
49
97
24
64
21
64
29
63
96
218
78
243
Post-mortem examination
The procedure for post-mortem examination was based on the guidelines suggested by Siccardi and Burmester (1970) and it was also consistent with recommendations of Purchase and Sharma (1973). AH birds dying during the test were examined for gross lesions. Briefly, birds with LL were always over 19 weeks of age and had tumours in the bursa of Fabricius. Some also had lesions in the viscera and muscles. Birds considered to have MD were of any age and had visceral, muscle and/or neural lesions but did not have tumours in the bursa of Fabricius. Because diagnoses were based on gross examination, it is possible that some bursal tumours due to LL were not noticed and that some of the bursal lesions were associated with MD (Payne and Rennie, 1970). Percentage mortality in the first five 28-d periods was based on the number of birds hatched. At the end of the fifth period, when the pullets were transferred into the laying houses, their number was slightly reduced but in the reduction no attention was paid to the health status of the birds. From the sixth 28-d period onward the mortality percentages were based on the number of birds housed in the laying house.
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
379
Statistical analyses
The all-or-none data on mortality in the meat strains tested were arranged into four-dimensional contingency tables of counts and were analysed by the method described by Fienberg (1970), using an iterative procedure to fit models by maximum likelihood. One of the variables was considered as a " response " factor always expressed as a dichotomy (e.g. for total mortality the response categories were " dead " or " alive " ) , while the remaining variables (hatch, vaccination and strain) were regarded as " design " variables since the marginal totals for these categories were predetermined before the experiments were initiated. Using the same notation as Fienberg (1970), the sequence of models tested started from model log em
= [i] + [ 4 + [B]j + [C]k +[D]i + [CD]kl + [BD]n + [BC]}k + [AD]U + [AC]ik + [AB]{j + [BCD]m + [ACD]m + [ABD]m + [ABC]m,
(1)
Downloaded by [New York University] at 16:04 11 April 2015
where: eijki = expected count in the ith hatch, jth category of vaccination, kth. strain and /th category of response, [1] = the grand mean of the logarithms of expected counts, [A]i, [B]}, [C]fc and [Z)]j = parameters representing deviations from the grand mean due to the hatch, vaccination, strain and response categories respectively, [CD]ici, [BD]}i,..., [ABC]i]h = two-factor and three-factor effects or interactions.
Subsequent models in the sequence were formed by omitting terms from equation (1) but retaining all parameters involving only " design " variables. Thus the simplest model that could be considered would be: log em
= [i] + [Ay + [B], + [C]k + [D]t + [AB]fi + [AC]ik + [BC]Jk + [ABC]m.
No term may be omitted from the model unless all higher-order interactions involving that term have also been omitted. The adequacy of a particular model was examined using a log-likelihood ratio statistic (LLR). By differencing the LLR values for a pair of successive models, one model generated from the other (the " reference model ") by dropping the parameters for a particular interaction, an LLR value corresponding to that source of variation was obtained. Provided the reference model had given an adequate fit to the data this LLR value was compared with x2 tables to obtain a test of significance. In this way, the LLR for equation (1) may be partitioned into components as in an analysis of variance. When a stage was reached at which no non-significant term could be dropped from the model and at which the model gave an adequate fit, that model was called the " final model ". It should be noted that the sequence of models used is not unique and, because of the sequential procedure, the test of significance for any component, or source of variation, after the first should be interpreted with the usual caution. The same general procedure could be used to assess the effect of strain on mortality in the vaccinated Leghorns and in the vaccinated and non-vaccinated meat-type birds separately. Instead, the analogous procedure of analysis of variance for categorical data in 2-way tables (Catanova) described by Light and Margolin (1971) was used. A feature of the Catanova analysis is the R 2 statistic, which
CO CO
Downloaded by [New York University] at 16:04 11 April 2015
o
TABLE 3
Dependence of mortality in meat-type females on hatch, vaccination and strain of chickens. Results of filing models to the data by maximum likelihood
Adult mortality
Rearing mortality Marek's disease
Total Source variation Hatch (H) Vaccination (V) Strain (S) HV HS VS HVS Final models: Sources retained LLR for goodness of fit Degrees of freedom 1
Degrees of freedom i i
7
i 7 7 7
LLR1 2-84 I4-753 3
5O-39 2-38
5-o6 15-78*
V, S, VS 23-59 48
SEQ? 4
LLR o-68 18-60** 19-73** o-oi 5-38 4-41 8-40 V,S 18-88 4'
SEQ. 4
LLR 0-42 47-51* 90-36* 0-26 12-61
SEQ. 4
18-45* 14-07 V, S, VS 27-36 48
Log-likelihood statistic corresponding to the dependence of mortality on the factors or interactions. Sequence of dropping the sources of variation from the complete model. 3 Relative to model from which the significant VS interaction was excluded. No test of significance made. * LLR statistic exceeds %" value for P = 0-05. ** LLR statistic exceeds x2 value for P = o-oi. 2
Lymphoid leukosis
Marek's disease
Total
LLR 1-36 63-34s 20763 o-8o
6-oo 15-69* 10-63 V, S, VS 18-79 48
SEQ,
LLR
4
0-02
o-54 62-66** 0-04 11-47 6-13
8-34 S 26-54 40
SEQ 4 6
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
381
measures the proportion of the total variation (in the sense defined by Light and Margolin, 1971) which is accounted for by knowledge of the strain. For large
Downloaded by [New York University] at 16:04 11 April 2015
NON-VACCINATED
.1
8
9
10
11
12
13
MORTALITY:
•
NONSPECIFIC
5J
14
LYMPHOID LEUKOSIS
WITHOUT NEURAL INVOLVEMENT
VACCINATED FOR MD
WITH NEURAL INVOLVEMENT
IU
2 cc o
II 2
3
4
5
6 7 8 9 28-DAY PERIODS
10
11
12
13
14
FIG. 1.—Mortality to 392 d in vaccinated and non-vaccinated meat-type birds, strains and hatches combined. H = housed in laying house.
samples, the log likelihood and Catanova procedures are approximately equivalent and the Catanova has been applied here to illustrate its use.
382
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
RESULTS
Meat-type females
Results of fitting models to the mortality data in meat-type females are shown in Table 3. The final models for total rearing and total adult mortality both contained parameters indicating influence of vaccination and strain of chickens on mortality, as well as interaction of vaccination'and strain with mortality. Hatch and interactions of hatch with vaccination or strain were shown to be unimportant MORTALITY:
•
NONSPECIFIC LYMPHOID LEUKOSIS
MAREK'S DISEASE
Downloaded by [New York University] at 16:04 11 April 2015
4-
WITHOUT NEURAL INVOLVEMENT WITH NEURAL INVOLVEMENT
LJ
8 3 UJ
2-
o
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
28-DAY PERIODS FIG. 2.—Mortality to 497 d in vaccinated Leghorns, all strains combined. H = housing in laying house.
not only in total mortality but also in MD and LL mortality (Table 3). Therefore, mortality from both hatches was combined in Figs. 1 and 3 and hatch was disregarded in the Catanova analyses (Table 4). Vaccination for MD reduced the total mortality in all meat strains combined from 11 % to 7% in the rearing period and from 32*4% to 20-1 % in the adult period. The importance of the effect of genotype was illustrated by the fact that during the rearing period total mortality in vaccinated birds varied amongst strains from 3*1% to 16-7% and in non-vaccinated birds from 4*9% to 18-7% (Fig. 3). In the adult period, when mortality was generally higher, strain differences in total mortality were even more pronounced. The lowest total mortality of adult birds (6*i% and 13*5% in vaccinated and non-vaccinated birds respectively) was observed in the unselected control strain K with the lowest average 56-d body weight. On the other hand, strain B with the third highest average 56-d body weight, had the highest mortality among both vaccinated (38-6%) and non-vaccinated (41*3%) birds (Fig. 3). The results of the analyses of variance for categorical data (Table 4) confirm the significance of the effects of strain on total mortality. However, the proportion
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
383
of variation in total mortality explained by strain was generally low and ranged from 1-5% in the non-vaccinated birds in the rearing period to 6-2% in vaccinated birds during the adult period (Table 4). MORTALITY:
•
NONSPECIFIC
40-
ED
LYMPHOID LEUKOSIS
0-140 DAYS OF AGE
£ 30
MAREK'S DISEASE
UJ
o cc 111
WITHOUT NEURAL INVOLVEMENT
LUI
WITH NEURAL INVOLVEMENT
EB
o.
Downloaded by [New York University] at 16:04 11 April 2015
> 20-
NV
Ioc
o
NV
10-
NV •
V
INV V NV
A (1347)
H (1281)
B (1195)
C (1165)
D (1111)
r
E (1098)
V V
f—inor F (1098)
K (873)
STRAIN (AVERAGE 56-DAY BODY WEIGHT IN GRAMS) 143-392 DAYS OF AGE
40-^
V NV
30' u cc
HI
t
£
20
cc
o 5
IOH
A (1347)
H (1281)
B (1195)
C
D
smi
E F K (1165) (1111) (1098) (1098) STRAIN (AVERAGE 56-DAY BODY WEIGHT IN GRAMS)
(873) ,
FIG. 3.—Mortality in meat-type strains. Strains ranked left to right on average 56-d body weight. V = vaccinated for MD, NV = non-vaccinated.
The fact that vaccination reduced the total mortality to different extent in the different strains is clearly seen especially in the adult period (Fig. 3). For example
384
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
the two strains (H and B) with the second and third highest average 56-d body weight showed little or no reduction in mortality with vaccination. Yet strain A, with the highest 56-d body weight, gave a reduction in total mortality from 2 5 7 % to 12-4%, a reduction of the same order as all the other strains. Differences among strains were greater for vaccinated than for non-vaccinated birds (Fig. 3) and this is reflected in the R 2 values shown in Table 4. Monthly mortality of all meat strains combined shown in Fig. 1 indicates that losses increased from the first to the seventh or eighth 28-day period followed by a slight decline. This trend is observed in both the vaccinated and non-vaccinated birds despite the generally lower mortality of the vaccinates. TABLE 4
Downloaded by [New York University] at 16:04 11 April 2015
Proportion of variation in mortality explained by strain 0/chickens (iJ2) and Catanova statistic (C) for differences between strains Meat-type females (df = 7) Vaccinated for MD
Non-vaccinated
Vaccinated Leghorn females (df = 9)
C
R 2 (%)
C
R 2 (%)
G
R 2 (%)
Rearing mortality: Total Marek's disease
47-8* 2-9
2-9 02
24-4** 20-0**
1-5 1-3
29-8** o
0-3 o
Adult mortality: Total Marek's disease Lymphoid leukosis
8I-I** 25-1** 47-8**
6-2 1-9 3-7
31-3** 20*3** 30-0**
2-6 1-3 2-5
23-1** 56-9** 54'3**
o-6 0-7 °'7
* P g 0-05; ** P g o-oi.
The analysis of data on rearing mortality due to MD indicated the dependence of mortality on vaccination and strain only, while the final model for MD mortality in the adult period also contained a parameter for vaccination by strain (by mortality) interaction (Table 3). Overall mortality attributed to MD was reduced by vaccination from 2*4% to o-6% in the rearing period and from 14-0% to 4-8% in the adult period. The Catanova analyses (Table 4) indicated that the significance of strain effects on rearing mortality attributed to MD resulted from strain differences among non-vaccinates. The strain effects on adult MD mortality were highly significant in both vaccinated and non-vaccinated birds (Table 4). While rearing MD mortality in non-vaccinated birds ranged from 5*1% for strain F to o*6% for strains B, G and D (Fig. 3), it was no greater than i-2% for any strain in vaccinated birds. In non-vaccinated adult birds MD mortality ranged from 20-1% for strain C to 9% for strain A (Fig. 3). Although all strains showed some decrease in MD mortality with vaccination, the reduction was 15% for strain C but less than 2% for strain H. In the non-vaccinated meat-type females MD mortality increased from the 3rd to the 7th month (Fig. 1). The first deaths due to MD occurred in nonvaccinated birds at 69 and 55 d of age in the first and second hatches respectively. In the corresponding vaccinated flock MD mortality was delayed respectively to 101 and 105 d in the first and second hatches and an increase was observed between the 4th and 8th month. In both the vaccinated and non-vaccinated flocks some MD mortality was observed to the end of the experiment.
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
385
Mortality from LL was found only in the adult period. It was influenced significantly by strain of chickens (Table 3) but there was no difference between MD vaccinated (3-4%) or non-vaccinated (2-9%) hens. The strain effect was highly significant in both vaccinated and non-vaccinated birds when analysed separately but the proportion of variation in LL mortality explained by strain was low (Table 4). The time pattern of mortality due to LL in meat-type females was not changed by vaccination (Fig. 1). The first LL mortality was observed at 141 d in the first hatch and at 135 d in the second hatch. MORTALITY: 40-1
LYMPHOID LEUKOSIS
0-140 DAYS OF AGE
Downloaded by [New York University] at 16:04 11 April 2015
30-
MAREK'S DISEASE
u cc U
a. >•
a
NONSPECIFIC
WITHOUT NEURAL INVOLVEMENT WITH NEURAL INVOLVEMENT
GD
•
20-
cc
o
10-
0J
•
1—1 t=3
1 (75)
4 9 8 (74) (73) (73)
c=t_
n
B3
«a
P
1 1
3 2 7 6 10 (73) (72) (71) (70) (69) STRAIN (AVERAGE HEN-DAY PERCENT PRODUCTION FROM FIRST EGG TO 497 DAYS)
5 (66)
40-, 141-497 DAYS OF AGE
30-
20cc
o s 10
1 (75)
3 2 7 6 10 (73) (72) (71) (70) (69) STRAIN (AVERAGE HEN-DAY PERCENT PRODUCTION FROM FIRST EGG TO 497 DAYS)
4 (74)
9 (73)
8 (73)
5 (66)
FIG. 4.—Mortality in vaccinated Leghorn strains. Strains ranked left to right on average hen-day percentage egg production from the first egg to 497 d.
386
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
Downloaded by [New York University] at 16:04 11 April 2015
Leghorn females
The mortality in the vaccinated Leghorn females (Fig. 4) was generally low in all strains. The total mortality to 497 d was 11 % including 2-1 % mortality from MD and 1-2% mortality from LL. Effects of strain were highly significant in all types of mortality examined except for MD mortality in the rearing period (Table 4). However, the R 2 values for the effect of strain, shown in the same table, are even lower than those calculated for the meat-type birds. In the adult period, the unselected control strains 5, 6 and 7 tended to have greater total and MD mortality than all the high egg production selected strains except strain 10 (Fig. 4). Also, control strain 5, which had the lowest egg production rate, had the highest MD mortality of all the strains tested. The monthly mortality pattern in the Leghorns (Fig. 2) was quite similar to that observed in vaccinated meat-type females. The first death from MD occurred in the Leghorns at 77 d of age while the first mortality from LL was observed at 162 d of age. DISCUSSION
The results of this study were generally consistent with previous findings. The repeatedly demonstrated reduction of MD mortality by vaccination (Purchase, 1973) was confirmed by the data on meat-type females. Vaccination delayed the first deaths of meat-type females from MD by 34 and 50 d in the first and second hatch respectively. This seems to support the observation by Spencer and Robertson (1974) that vaccination blocked infection with virulent MD virus for approximately 5 weeks after contact exposure. The importance of genetic resistance for successful vaccination (Spencer etal., 1972, 1974) was clearly indicated in this study. However, contrary to another report (Spencer et al., 1975), vaccination did not seem to reduce the strain variation in MD mortality (Table 4). The proportion of variation in mortality explained by strain was low although the effects of strain on mortality were, in most instances, -highly significant. This illustrates the fact noted by Light and Margolin (1971) that, in large sets of data, weakly related variables can nevertheless exhibit statistically significant dependencies. The data in Fig. 4 suggest that the viability of the Leghorn strains selected for both egg production and viability was higher than in the unselected control strains. These results support previous reports (Von Krosigk et al., 1972; Gavora et al., 1974) that a positive genetic correlation exists between resistance to MD and egg production. Examination of mortality and growth rate of the meat strains in Fig. 3 suggests a negative correlation. Of the two unselected control strains (Table 2) only strain K remained unselected since 1958. The other control strain F had been synthesised in 1963 from strains selected for rapid growth since 1958. The 56-d body weight, as well as mortality of the control strain F were similar to those of the growth-selected strains. On the other hand, the control strain K with a substantially lower 56-d body weight than the selected strains showed the lowest mortality of the vaccinated birds in the rearing period and ranked third best among non-vaccinates (Fig. 3). In the adult period strain K had the lowest total mortality. In MD mortality strain K ranked as the second lowest among vaccinates and as the lowest among non-vaccinates. This supports reports of Friars et al. (1972), Han and Smyth
MAREK'S DISEASE AND LYMPHOID LEUKOSIS
387
Downloaded by [New York University] at 16:04 11 April 2015
(1972) and Gavora et al. (1974) indicating a negative genetic correlation between growth rate and resistance to Marek's disease. However, strain A was an exception since it had the highest 56-d body weight and was relatively resistant to MD (Fig. 3). Thus, it would appear possible to produce rapidly growing strains of chickens with good general viability and resistance to MD and LL. Of special interest are the results of this study on the relative importance of MD and LL in vaccinated meat-type breeders and in vaccinated Leghorns. In chickens vaccinated for MD the total lifetime mortality was 11% in Leghorns and 27% in meat-type females. Approximately one-fifth of this mortality was caused by MD and one-tenth by LL. Hence, losses from MD and LL combined amounted to onethird of total lifetime mortality in vaccinated Leghorn and meat-type hens even with the use of an effective MD vaccine. The continuation of research work aimed at further reduction of these losses would seem desirable. ACKNOWLEDGEMENTS
We gratefully acknowledge the technical assistance of Mr P. Apedaile who carried out the post-mortem examinations. We also wish to thank Mr J . Dickie for his valuable technical assistance and Miss L. Francis, Statistical Research Service, Agriculture Canada, for carrying out the maximum likelihood analyses. The suggestions of Mrs P. Morse from the Statistical Research Service and of Dr J . L. Spencer from the Animal Diseases Research Institute, Hull in Ottawa in preparation of the manuscript are gratefully acknowledged. REFERENCES
FIENBERG, S. E. (1970). The analysis of multidimensional contingency tables. Ecology, 5 1 : 419-432. FRIARS, G. W., CHAMBERS, J . R., KENNEDY, A., AND SMITH, A. D. (1972). Selection for resistance to
Marek's disease in conjunction with other economic traits in chickens. Avian Dis., 16: 2-10.
GAVORA, J . S., GRUNDER, A. A., SPENCER, J . L., GOWE, R. S., ROBERTSON, A. AND SPECKMANN,
G. W. (1974). An assessment of effects of vaccination on genetic resistance to Marek's disease. Poult. Sci., 53: 889-897.
GOWE, R. S., LENTZ, W. E. AND STRAIN, J . H. (1973). Long-term selection for egg production in
several strains of White Leghorns: Performance of selected and control strains including genetic parameters of two control strains. 4th Europ. Poult. Conf., London, 225-245. HAN, F. S. AND SMYTH, J . R. (1972). The influence of growth rate on the development of Marek's disease in chickens. Poult. Sci., 5 1 : 975-985. HAN, F. S., SMYTH, J . R., SEVOIAN, M. AND DICKINSON, F. N. (1969). Genetic resistance to leukosis
caused by the J M virus in the fowl. Poult. Sci., 48: 76-87. KING, S. C., CARSON, J. R. AND DOOLITTLE, D. P. (1959). The Connecticut and Cornell random bred populations of chickens. Wld's Poult. Sci. J., 15: 139-159. LIGHT, R. J . AND MARGOLIN, B. M. (1971). An analysis of variance for categorical data. J. Am. statist. Ass., 66: 534-544. MERRITT, E. S. AND GOWE, R. S. (1962). Development and genetic properties of a control strain of meat-type fowl. XIth Wld's Poult. Congr., Mexico, 66-70. PAYNE, L. N. AND RENNIE, M. (1970). Lack of effect of bursectomy on Marek's disease. J. natn. Cancer. Inst., 4 5 : 387-391. PURCHASE, H. G. (1973). Control of Marek's disease by vaccination. Wld's Poult. Sci. J., 29: 238-250. PURCHASE, H. G., OKAZAKI, W. AND BURMESTER, B. R. (1971). Field trials with the herpes virus of turkeys (HVT) strain FC126 as a vaccine against Marek's disease. Poult. Sci., 50: 775-783.
388
J. S. GAVORA, E. S. MERRITT, A. A. GRUNDER AND R. S. GOWE
PURCHASE, H . G., OKAZAKI, W. AND BURMESTER, B. R. (1972). Long-term field trials with the herpesvirus of turkeys vaccine against Marek's disease. Avian Dis., 16: 57-71. PURCHASE, H . G. AND SHARMA, J . M. (1973). The differential diagnosis of lymphoid leukosis and Marek's disease. Am. Ass. Avian Pathologists, Slide Study Set 3. SICCARDI, F. J . AND BURMESTER, B. R. (1970). The differential diagnosis of lymphoid leukosis and Marek's disease. Tech. Bull. 1412, Washington, Agricultural Research Service, United States Department of Agriculture. SPENCER, J . L., GAVORA, J . S., GRUNDER, A. A., ROBERTSON, A. AND SPECKMANN, G. W. (1974).
Immunization against Marek's disease: Influence of strain of chickens, maternal antibody and type of vaccine. Avian Dis., 18: 33-44. SPENCER, J . L., GRUNDER, A. A., ROBERTSON, A. AND SPECKMANN, G. W. (1972).
Attenuated
Marek's disease herpesvirus: protection conferred to strains of chickens varying in genetic resistance. Avian Dis., 16: 94-107. SPENCER, J . L., AND ROBERTSON, A. (1975). Influence of vaccination with avirulent herpesviruses on subsequent infection of chickens with virulent Marek's disease herpesvirus. Am. J. vet. Res. (in press).
Downloaded by [New York University] at 16:04 11 April 2015
VON KROSIGK, C. M., MCCLARY, C. F., VIELITZ, E. AND ZANDER, D. V. (1972).
Selection for
resistance to Marek's disease and its expected effects on other important traits in White Leghorn crosses. Avian Dis., 16: 11-19.
