ArticlePDF Available

Modelling of Rhode Island Red chicken strains

Authors:
HINA KAUSAR
HINA KAUSAR
HINA KAUSAR
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Med Ram Verma at Indian Agricultural Statistics Research Institute
Med Ram Verma
Sanjeev Kumar at Bihar Animal Sciences University Patna
Sanjeev Kumar
  • Bihar Animal Sciences University Patna
Vijay B. Sharma at Indian Veterinary Research Institute
Vijay B. Sharma
Show all 6 authorsHide
Download full-text PDF
Read full-text
Download citation

Abstract and Figures

To study the growth pattern in body weight of 3 strains of Rhode Island Red chicken Bertalanffy, gompertz and logistic nonlinear models were fitted. From the data on body weights of three strains of Rhode Island Red, we observed that average body weights of male chicken were higher than the female chicken. Based on the various measures of goodness fit criteria we have observed that in modelling of body weight of the Rhode Island Red chicken Bertalanffy was the best fitted model. In case of Rhode Island Control, Bertalanffy was the best fitted model and for Rhode Island Control male chicken logistic was the best fitted model. In case of Rhode Island White chicken logistic was the best fitted model and in case of Rhode Island White male chicken Bertalanffy was the best fitted model. In case of female chicken of Rhode Island Red, Rhode Island Control and Rhode Island White strains gompertz model was the best fitted model. From these fitted models one can determine the expected average body weight of a group of birds of three strains of RIR chicken at any given age under normal conditions. © 2016, Indian Council of Agricultural Research. All rights reserved.
Figures - uploaded by Ananta Kumar Das
Author content
All figure content in this area was uploaded by Ananta Kumar Das
Content may be subject to copyright.
Content uploaded by Ananta Kumar Das
Author content
All content in this area was uploaded by Ananta Kumar Das on Aug 24, 2016
Content may be subject to copyright.
120
Present address: 1,6M.V.Sc. Scholar (sanavet13@gmail.com,
dilliwar123@gmail.com), 2Senior Scientist (medramverma
@rediffmail.com), 4Ph.D. Scholar (drvijaybhardwaj
@gmail.com), Division of Livestock Economics, Statistics and
Information Technology. 3Principal Scientist (skgicar
@gmail.com), ICAR- Central Avian Research Institute, Izatnagar.
5Subject Matter Specialist (dasugenvet@gmail.com), Howrah
KVK, West Bengal.
Poultry farming has assumed much importance due to
the growing demand of poultry products especially in urban
areas because of their high food value. The poultry has the
highest rate of growth in agricultural sector in India with a
growth rate of 8 to10% in eggs and 15 to 20% in broilers
over the last two decades. From the year 1999–2000
onwards the production of egg improved substantially and
it reached 69,731 million numbers in the year 2012–13
(BAHS 2014). The growth is an irreversible, correlated and
coordinated increase in the mass of body in a definite
interval of time (Brody 1945). It is necessary to have
knowledge of pattern of growth in poultry because body
growth is an important factor that contributes to the
profitability in poultry production. Fitting growth curves
to longitudinal measurements is a standard method to
analyze growth pattern in poultry (Berkey 1986). Growth
models relate the average weight of different poultry birds
as a function of their age. From these models one can
determine the expected average weight of a group of birds
of the same breed at any given age under normal conditions
(Prasad and Singh 2006). Several nonlinear models are
being used in poultry which define the relationship between
age and live weight (Prasad et al. 2008, Prasad and Singh
2006, Paul et al. 2011). The objective of these growth curves
is to describe the behaviour of the body weight with increase
in time or age with mathematical parameters that are
biological interpretable.
Data on body weights of 3 pure strains of Rhode Island
Red were compiled from the study conducted by Das (2013)
during 2011 at Central Avian Research Institute, Izatnagar,
Bareilly, Uttar Pradesh. For analysing the growth pattern
of body weight of Rhode Island Red strains chicken
Indian Journal of Animal Sciences 86 (5): 612–615, May 2016/Short communication
Modelling of Rhode Island Red chicken strains
HINA KAUSAR1, MED RAM VERMA2, SANJEEV KUMAR3, VIJAY BAHADUR SHARMA4,
ANANTA KUMAR DAS5 and LEENA DILLIWAR6
ICAR - Indian Veterinary Research Institute, Izatnagar, Bareilly, Uttar Pradesh 243 122 India
Received: 29 May 2015; Accepted: 28 September 2015
Key words: Adjusted R2, Durbin Watson (DW) test, MAE, MSE, Non linear models, R2
(Graphical abstract available only online)
following nonlinear models were used (Prasad et al. 2008,
Singh et al. 2014).
Gompertz model: yt = a exp (–b exp (–ct)+e
Logistic model:
Bertalanffy model: yt = a [1-b exp (-ct)]3 +e
where yt, observed body weight of the chick at tth week; t,
age of the chick in weeks; e, the error term; a, asymptotic
weight; b, scaling parameter; c, rate of maturity.
The goodness of fit of the models was checked by
coefficient of determination (R2), adjusted coefficient of
determination ( 2), Mean square error (MSE), mean
absolute error (MAE) and Akaike Information
Criterion (AIC). To test the assumptions about the
independence of errors Durbin Watson (DW) test was used
and Shapiro Wilk’s test was used to test the normality of
errors. For studying the growth pattern in poultry, data on
average body weights of 3 strains of Rhode Island Red
namely Rhode Island Red (1,308 birds), Rhode Island
Control (643 birds) and Rhode Island White (232 birds)
were used. The data on average body weights of chicken
during 0, 1, 2, 3, 4, 6, 8 and 12 weeks of 3 Rhode Island
Land strains are given in Table 1.
The average body weight of the male chicken varied
between 37.64 g in 0 week to 1,050.42 g in 12th week.
However the body weight of the female chicken varied
between 37.32 g in 0 week to 914.61 g in 12th week
(Table 1). The estimates of the growth parameters of Rhode
Island Red chicken strains are given in Table 2. The R2
values were high for all growth models indicating a
significant relationship between age and weight in both
sexes for Rhode Island Red chicken. Based on the different
goodness of fit criteria the Bertalanffy was the best fitted
model with the highest values of R2, adjusted R2 (2
R) and
lowest values of MAE and AIC for Rhode Island Red
chicken. For the male birds of RIR, Bertalanffy gave the
best fit and for the female chicken gompertz was the best
fitted model on the basis of different goodness of fit criteria.
May 2016] MODELLING OF RHODE ISLAND RED CHICKEN STRAINS 613
121
Durbin Watson (DW) test indicated that there was no
autocorrelation and Shapiro-Wilk’s test indicated that errors
were normally distributed. However, Prasad et al. (2008)
also observed that gompertz model best described the
growth pattern in Indian native chicken. Kuhi et al. (2003)
also observed the better performance of Bertalanffy model
than logistic and gompertz models for describing the growth
performance of chicken.
Rhode Island control male and female chicken varied
from 34.35 g during 0th week to 765.43 g in 12th week. The
R2 values were high for all growth models indicating a
significant relationship (Table 3) between age and weight
in both sexes for Rhode Island Control chicken. Based on
the various goodness of fit measures Bertalanffy was the
best fitted model for Rhode Island Control chicken strain.
For the body weights of the male RIC logistic model best
described the data with highest R2 and adjusted R2 and
minimum MSE, MAE and AIC. For the female birds of
RIC strain Gompertz model gave the best fit with highest
R2 and adjusted R2 and minimum value of MSE, MAE,
and. Durbin Watson (DW) test indicated that there was no
autocorrelation and Shapiro-Wilk’s test indicated that errors
were normally distributed. The gompertz function has been
preferred over the logistic function for fitting monophasic
growth curves of chickens (Laird 1966). However Prasad
et al. (2008) also observed that gompertz model best
described the growth pattern in Indian native chicken. Kuhi
et al. (2003) also observed the better performance of
Bertalanffy model than logistic and gompertz models for
describing the growth performance of chicken.
The estimates of the growth parameters of Rhode Island
White strain are given in Table 4. The R2 values were high
Table 1. Average body weight of different Rhode Island Red strains chicken during different weeks
Strain Sex Number Average body weight during different weeks
of birds 0123 4 6812
RIR Male 701 37.64 57.03 92.63 158.64 206.93 365.43 600.14 1050.42
Female 607 37.32 56.85 90.39 151.03 193.78 331.84 561.69 914.61
Combined 1308 37.90 56.95 91.56 155.08 200.71 350.56 585.41 985.55
RIC Male 355 34.54 52.49 82.37 144.56 174.37 279.77 422.13 765.43
Female 288 34.35 51.21 82.88 134.82 167.78 261.11 397.92 705.46
Combined 643 34.46 51.84 82.60 139.65 171.41 270.60 410.59 738.18
RIW Male 141 35.03 50.84 79.49 137.40 188.02 329.48 498.67 945.28
Female 91 35.35 50.59 76.82 134.63 181.01 311.04 488.18 811.01
Combined 232 31.16 50.75 78.55 136.46 185.32 322.10 479.37 892.30
Table 2. Parameter estimates of models and goodness of fit statistics of Rhode Island Red chicken
Sex Model Parameter Estimate SE R2 2 MSE MAE AIC DW
Male Bertalanffy a 6563.642 1086.229 0.999 0.998 200.4 9.56 53.7 2.13
b 0.832 0.006
c 0.05 0.006
Gompertz a 1754.504 663.234 0.882 0.811 376.63 12.29 59.38 2.51
b 19.903 33.947
c –0.302 0.214
Logistic a 2079.418 169.504 0.996 0.993 1574.465 27.098 72.255 2.62
b 28.194 3.62
c 0.287 0.023
Female Bertalanffy a 2443.846 264.176 0.999 0.998 660.026 11.577 64.43 2.82
b 0.794 0.015
c 0.087 0.009
Gompertz a 1829.161 95.666 0.999 0.998 229.7264 9.488 54.932 2.19
b 3.997 0.11
c 0.148 0.009
Combined Bertalanffy a 3849.163 463.684 0.999 0.998 214.084 9.58 54.297 2.74
b 0.81 0.008
c 0.067 0.006
Gompertz a 2478.382 173.105 0.999 0.998 40541.23 105.49 101.49 1.613
b 4.143 0.092
c 0.127 0.008
Logistic a 1720.349 110.931 0.996 0.994 1250.596 23.61 70.182 2.042
b 25.024 3.296
c 0.302 0.024
614 KAUSAR ET AL. [Indian Journal of Animal Sciences 86 (5)
122
Table 4. Parameter estimates of models and goodness of fit statistics of Rhode Island White
Sex Model Parameter Estimate SE R2 2 MSE MAE AIC DW
Male Bertalanffy a 878.483 158.527 0.917 0.867 11447.37 65.64 90.11 2.68
b 1.032 0.441
c 0.244 0.11
Gompertz a 856.487 123.208 0.926 0.809 12302.38 59.456 90.757 2.17
b 5.126 2.598
c 0.306 0.115
Logistic a 591.693 86.832 0.854 0.766 7906.561 48.148 86.779 3.12
b 19.969 22.328
c 0.52 0.229
Female Bertalanffy a 754.235 127.402 0.922 0.875 7790.591 53.541 86.64605 2.53
b 0.945 0.348
c 0.241 0.102
Gompertz a 736.171 100.047 0.93 0.906 6976.908 48.38913385.65325 2.72
b 4.524 1.966
c 0.3 0.106
Combined Bertalanffy a 828.165 146.05 0.918 0.869 9968.862 61.0941 88.865 2.57
b 0.999 0.406
c 0.244 0.107
Gompertz a 808.141 114.266 0.927 0.883 8933.686 55.2718 87.87 2.68
b 4.884 2.344
c 0.304 0.112
Logistic a 781.965 74.097 0.943 0.908 6946.761 44.62781 85.614 2.86
b 27.667 21.504
c 0.508 0.142
Table 3. Parameter estimates of models and goodness of fit statistics Rhode Island Control
Sex Model Parameter Estimate SE R2 2 MSE MAE AIC DW
Male Bertalanffy a 605.013 131.884 0.826 0.721 13270.28 65.18 91.44 2.15
b 0.931 0.542
c 0.278 0.171
Gompertz a 598.823 113.29 0.836 0.737 12522.16 61.27 90.91 2.34
b 4.307 2.95
c 0.333 0.181
Logistic a 591.693 86.832 0.854 0.766 11099.84 52.804 89.83 2.02
b 19.969 22.328
c 0.52 0.229
Female Bertalanffy a 568.367 121.035 0.837 0.739 10589.73 57.88819 89.408 2.06
b 0.89 0.47
c 0.269 0.159
Gompertz a 561.344 102.636 0.847 0.769 9943.89 54.52708 88.84242 2.19
b 4.113 2.577
c 0.325 0.168
Combined Bertalanffy a 3559.83 473.48 1 1 83.67 6.025 45.84 2.25
b 0.793 0.006
c 0.055 0.005
Gompertz a 2050.16 126.87 1 1 40351.58 105.57 101.44 3.2
b 3.94 0.052
c 0.112 0.006
Logistic a 1318.23 68.58 0.998 0.996 348.84 14.25 58.69 0.25
b 22.088 1.852
c 0.282 0.016
May 2016] MODELLING OF RHODE ISLAND RED CHICKEN STRAINS 615
123
for all growth models indicating a significant relationship
between age and weight in both sexes for Rhode Island
White chicken. Based on the goodness of fit criteria logistic
model presented best adjustment to the body growth data
with maximum R2 and adjusted R2 and minimum values
of MSE, MAE and AIC. In case of male birds of the Rhode
Island White Bertalanffy gave the best fit as evident from
the highest value of R2, adjusted R2 and minimum values
of MSE, MAE and AIC. For the female birds gompertz
model gave the best fit with highest R2 and adjusted R2 and
lowest MSE and MAE. Durbin Watson (DW) test indicated
that there was no autocorrelation and Shapiro-Wilk’s test
indicated that errors were normally distributed. However,
Prasad and Singh (2006) observed that modified logistic
model best described the growth pattern in male and female
chicken. However, Prasad et al. (2008) observed
that gompertz model best described the growth pattern in
Indian native chicken. Kuhi et al. (2003) also observed the
better performance of Bertalanffy model than logistic and
gompertz models for describing the growth performance
of chicken.
SUMMARY
To study the growth pattern in body weight of 3 strains
of Rhode Island Red chicken Bertalanffy, gompertz and
logistic nonlinear models were fitted. From the data on body
weights of three strains of Rhode Island Red, we observed
that average body weights of male chicken were higher than
the female chicken. Based on the various measures of
goodness fit criteria we have observed that in modelling of
body weight of the Rhode Island Red chicken Bertalanffy
was the best fitted model. In case of Rhode Island Control,
Bertalanffy was the best fitted model and for Rhode Island
Control male chicken logistic was the best fitted model. In
case of Rhode Island White chicken logistic was the best
fitted model and in case of Rhode Island White male chicken
Bertalanffy was the best fitted model. In case of female
chicken of Rhode Island Red, Rhode Island Control and
Rhode Island White strains gompertz model was the best
fitted model. From these fitted models one can determine
the expected average body weight of a group of birds of
three strains of RIR chicken at any given age under normal
conditions.
ACKNOWLEDGEMENT
The authors are highly thankful to the learned referees
and the Assistant Editor for their valuable comments on
the original version of the paper.
REFERENCES
BAHS (Basic Animal Husbandry Statistics). 2014. Department
of Animal Husbandry, Dairying and Fisheries, Govt. of India.
Berkey C S. 1986. Nonlinear growth curve analysis: Estimating
the population parameters. Annals of Human Biology 13: 111–
28.
Brody S. 1945. Bioenergetics and Growth; With Special Reference
to the Efficiency Complex in Domestic Animals. pp: 491–661.
Hafner Press, Reinhold, New York, USA.
Das A K. 2013. ‘Microsatellite polymorphism, immunocompe-
tence profile and performance evaluation in Rhode Island Red
chicken and its crosses.’ Ph.D. thesis, Indian Veterinary
Research Institute, Izatnagar.
Kuhi H D, Kebreab E, Lopez S and France J. 2003. An evaluation
of different growth function for describing the profile of live
weight with time (age) in meat and egg strains of chicken.
Poultry Science 82: 1536–43.
Laird A K. 1966. Postnatal growth of birds and mammals. Growth
30: 349–63.
Paul A K, Singh S, Kumar A, Singh O, Raman R K, Haunshi S
and Verma M R . 2011. Nonlinear growth models for body
growth of Vanraja poultry birds. IUP Journal of Genetics and
Evolution 4 (4): 65–69.
Prasad Shiv and Singh D P. 2006.An adjustment model of Logistic
form to describe the growth pattern of chickens. Indian Journal
of Poultry Science 41 (3): 280–82.
Prasad Shiv, Singh D P and Singh Rajendra. 2008. Nonlinear
describe growth pattern of Indian native chickens. Indian
Journal of Animal Sciences 78 (6): 645–48.
Sengul T and Kiraz S. 2005. Nonlinear model for growth curve
in large white turkeys. Turkish Journal of Veterinary and
Animal Science 29: 331–37.
Singh Umesh, Singh R B and Gautam S S. 2014. Nonlinear
stochastic model for describing growth of soybean production
in Madhya Pradesh and India. International Journal of
Agricultural and Statistical Sciences 10 (2): 375–79.
616 KAUSAR ET AL. [Indian Journal of Animal Sciences 86 (5)
124
... In poultry, body weight is one of the most significant traits and is associated with the bird's growth rate (Kumar et al., 2022). Growth occurs as a result of a continual, correlated, and coordinated rise in body weight or body mass over a certain length of time (Kausar et al., 2016;Kumar et al., 2022). Growth of the birds up to the point of marketing age is the most crucial, for which, body weights at different time points may be recorded and analyzed. ...
... Depending on the goal, growth patterns can be examined throughout a variety of time frames and can be explained using mathematical equations fitted to a growth curve (Kumar et al., 2022). The aim of the growth curve is to describe the rise in body weight over time or age using mathematical parameters that are biologically interpretable (Brody, 1945;Kausar et al., 2016;Kumar et al., 2022). Statisticians use different models for assessment of the best fitted model for growth analysis. ...
... A nonlinear model, which would be unbiased, offers a better fit since it generates less residuals (Kumar et al., 2022). Several nonlinear models have been used in poultry which could define the relationship between age and live body weight (Paul et al., 2011;Kausar et al., 2016). From these models, one can determine the expected average body weight of a group of birds of the same breed at any given age under normal conditions (Kausar et al., 2016). ...
Analysis of growth pattern of Rhode Island Red chicken using nonlinear models
Article
Full-text available
  • Dec 2022
Growth traits are quantitative in nature, with a continuum between high and low performing birds. Growth of the birds up to the point of marketing age is the most crucial, for which, body weights at different time points may be recorded and analyzed. This study was aimed to model the body weights of Rhode Island Red (RIR) chicken for assessment of the best fitted nonlinear model for growth analysis in the present data set. The data on body weights at different ages of RIR chicken were recorded from the experimental layer farm of the ICAR-Central Avian Research Institute, Izatnagar (India)and sex-wise average body weights were estimated. Different nonlinear statistical models (Gompertz, Logistic and Bertalanffy models) were fitted to the body weights, and the goodness of fit of the models was tested using different test statistics. To test the assumptions about the independence of errors in the models, Durbin Watson (DW) test was used, and to test the normality of the errors, Shapiro Wilk’s test was used. It was observed that the logistic model was the best fitted model for describing the variations in the body weights of male chicks, while Gompertz model was found to be the most suitable model for explaining growth patterns in female chicken as well as in combined sexes. The expected average body weight of RIR chicken could be determined at any given age under normal conditions utilizing these fitted models.
View
... En contraste, Galeano-Vasco y Cerón-Muñoz (2013), reportaron en pollas Lohmann LSL, hasta 196 días de edad, que el modelo Von Bertalanffy fue el que mejor ajustó los datos de crecimiento. Respecto al modelaje del crecimiento de aves Rhode Island Red, Kausar et al. (2016) encontraron en aves de 0 a 84 días de edad, que con el modelo Gompertz se obtenía un mejor ajuste de los datos. Silva et al. (2016), utilizaron el modelo Gompertz para estudiar el crecimiento de hembras Hy-Line Brown, de 0 a 126 d de edad. ...
... Estas diferencias pueden explicarse por la naturaleza de las poblaciones de aves CMX de cada estudio, así como las condiciones de manejo y ambientales. Los resultados obtenidos con aves RIR, contrastan con Kausar et al. (2016), quienes estimaron en aves RIR, de 0 a 84 días de edad, un peso asintótico superior (2 443.8 g) al del presente caso con el MNVB y un valor inferior (1 829.2 g) con el modelo Gompertz. Pero, las tasas de maduración correspondientes fueron superiores (0.087 y 0.148 día −1 ), en ambos modelos en comparación con los estimados en la presente investigación. ...
... Pero, las tasas de maduración correspondientes fueron superiores (0.087 y 0.148 día −1 ), en ambos modelos en comparación con los estimados en la presente investigación. De acuerdo a los criterios de bondad de ajuste (AR 2 y AIC), considerados por Kausar et al. (2016), concluyeron que con el modelo Gompertz se obtiene un mejor ajuste de los datos, hecho que difiere de los presentes resulta-dos, que indican un mejor ajuste por parte del MNVB (Tabla 2). Al respecto, y Grossman y Bohren (1985) estimaron en hembras RIR, con el MNL, edades al punto de inflexión (105.7 y 109.2 días de edad, respectivamente) superiores a los observados en este estudio. ...
Modelación del crecimiento de aves Criollas Mexicanas, Rhode Island Red y Hy-Line Brown
Article
Full-text available
  • Oct 2022
El objetivo fue caracterizar el crecimiento de aves Criollas Mexicanas (CMX), Hy-Line Brown (HLB) y Rhode Island Red (RIR) mediante los modelos Gompertz-Laird (MNGL), Logístico (MNL), Richards (MNR) y Von Bertalanffy(MNVB). Se utilizaron 70 pollitas de cada genotipo. Cada ave fue identificada para registrar su peso de 7 a 210 d de edad. Los parámetros de los modelos se estimaron con el procedimiento NLIN (algoritmo de Marquardt) de SAS. Se calculó el coeficiente de determinación ajustado, criterio de información de Akaike y criterio de información Bayesiana, para cada caso. La tasa de crecimiento inicial (d−1) en aves CMX fue menor (0.047 15) que en aves RIR (0.071 69) y HLB (0.052 16). Con los cuatro modelos, la tasa de decaimiento en aves CMX fue menor, respecto a los otros genotipos. Asimismo, la edad al máximo crecimiento (d) en las aves CMX,RIR y HLB, varió de 49.4 a 87.9, de 60.7 a 81.6 y de 58.7 a 87.8, respectivamente. Por otro lado, el peso asintótico (g), estimado con los cuatro modelos varió de 2 068.5 a 2 644.3, de 1 902.4 a 2 250.9 y 2 096.3 a 2 464.5, en las aves CMX, RIR y HLB, respectivamente. Con el MNVB se obtuvo el mejor ajuste para los datos de las aves RIR y HLB, en contraste, el MNR fue mejor para las aves CMX, seguido del MNVB. El MNVB sería el más adecuado para el estudio simultáneo del crecimiento de aves CMX, RIR y HLB.
View
... Rhode Island White [19]. ...
... Results of this study are in fully agreement with those reported previously where Logistic model was found to be the best fit growth model in females of Nicobari and Ghagus indigenous breeds of chicken [20]. Contrarily in females of selected flock, Gompertz [19,21]. ...
Growth Modelling in Aseel Native Chicken
Article
Full-text available
  • Jul 2022
Aseel, an important Indian native chicken breed, characterized by its aggressive behaviour, fighting ability and majestic gait is being utilized in India for developing crosses for backyard poultry farming. Therefore, information regarding its growth models is a pre-requisite which is yet not available. So, present investigation is first of its kind in India which was carried out to evaluate various non-linear models and to find out the best fitting growth model in Aseel. Data on body weights from second week of age to 20 weeks of age at biweekly intervals were recorded on 1054 progenies from 30 sires and 139 dams hatched in three hatches and maintained at Desi Fowl Unit of Central Avian Research Institute, Izatnagar, Bareilly. Owing to the non-linear characteristic of growth, three non-linear models namely, Gompertz model, Bertalanffy model and Logistic model were evaluated. Goodness of fit for all the models were checked using coefficient of determination (R2 ), adjusted coefficient of determination (Adj-R2 ), Mean Square Error (MSE), Mean Absolute Error (MAE) and Akaike Information Criterion (AIC). The body weights at 0, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 weeks of age averaged 28.18g, 65.94g, 138.99g, 237.06g, 359.95g, 533.74g, 718.28g, 922.33g, 1149.7g, 1321.65g and 1477.14g, respectively in males of Aseel native chicken. The corresponding body weights in females were 27.94g, 64.04g, 128.28g, 214.50g, 317.00g, 462.54g, 615.53g, 782.14g, 956.48g, 1077.65g and 1183.50g, respectively. Out of the three models, the Gompertz model, with R2 = 1.000, Adj�R2 = 1.000 and AIC = 29.839, best described the growth trend in males. In females, the Logistic model, with R2 = 0.999, Adj-R2 = 0.999 and AIC = 29.734 was found to be the best. In pooled or combined sex data, Gompertz model, having R2 = 0.999, Adj-R2 = 0.999 and AIC = 30.004, was determined as the best. The study revealed that these models can be used to determine the average body weights at any given point of time in Aseel native chicken under normal conditions. The investigation has generated baseline data on growth modelling and can also be used in other native chicken breeds. Keywords: Aseel; Growth Models; Gompertz Model; Logistic Model; Bertalanffy Model
View
... Besides more egg production, optimum growth in body weight is also an important attribute to the farmers for promoting RIR in rural livelihood. Notably, the growth is an irreversible, correlated and coordinated increase in the mass of body in a definite interval of time (Kausar et al. 2016). It is necessary to have knowledge of factors influencing the growth of poultry birds because body growth is an important factor that contributes to the profitability in poultry production (Kausar et al. 2016). ...
... Notably, the growth is an irreversible, correlated and coordinated increase in the mass of body in a definite interval of time (Kausar et al. 2016). It is necessary to have knowledge of factors influencing the growth of poultry birds because body growth is an important factor that contributes to the profitability in poultry production (Kausar et al. 2016). Therefore, the present study aimed for genetic analysis of grower body weights investigating the genetic and non-genetic parameters in a selected line of RIR chicken. ...
Genetic analysis of body weight in a selected line of Rhode Island Red grower chicken
Article
Full-text available
  • Mar 2019
A total of 1207 pedigreed progenies of 50 sires and 177 dams of a selected line of Rhode Island Red chicken were investigated in two hatches at ICAR-Central Avian Research Institute for genetic analysis of their body weights. The genetic and non-genetic parameters of chick weight at day-old age and body weight at different ages were estimated using least squares analysis of variance. The heritability and correlation coefficients were estimated for different traits using paternal half-sibs. The mean sum of squares for sex component of variance was significant for body weights at third week of age onwards, males being heavier throughout the ages. Sire component of variance was also significant for chick weight and different body weights except the estimates at 6 th and 16 th week of age. Hatch component of variance was highly significant for chick weight and body weights upto 12 th week of age. Regression effect of chick weight on subsequent body weights was found significant specifically for body weights from 1 st week onwards to 20 th week of age. The heritability estimates were 0.775±0.151 for chick weight, 0.303±0.09 to 0.44±0.142 for body weights from 1 st week to 12 th week of age, and 0.258 to 0.248 for body weights from 16 th week to 20 th week of age. The estimates at lower age could be used in selection programme based on the flock's own performance to improve those traits. The genetic and phenotypic correlation coefficients were positive and highly significant in most of the cases. The corresponding estimates ranged from 0.044 to 0.990 and 0.020 to 0.788, and these coefficients could be combined in construct of standard selection indices for optimum growth in body weight that might be adopted in future breeding strategy for this RIR chicken genotype.
View
... Contrastingly, reported that Gompertz model is the best model for describing growth in males, and Logistic model and Gompertz model for females in selected Aseel chicken. Kausar et al. (2016) reported Logistic model to be the best fit model in males and Gompertz model in females of random bred flock of Rhode Island Red and Rhode Island White chicken. They further reported that Bertalanffy model was the best model followed by Logistic model in Rhode Island Red and Rhode Island White random bred groups, which is quite similar to the present findings. ...
Bertalanffy Model Reflects Growth Trajectory in Aseel Chicken
Article
Full-text available
  • Nov 2022
Aseel, a popular breed of native chicken, characterized by its pugnacity, fighting strength and royal gait, is being used to create crosses for domestic chicken production. However, information on its growth models is scanty. An experiment was conducted to evaluate different non-linear models and to find out best fitting model in Aseel, being maintained at Central Avian Research Institute, Izatnagar, Bareilly. Data on body weights from 12-weeks of age to 20-weeks of age at biweekly intervals were recorded on a random bred single hatched flock. Owing to the non-linear characteristic of growth, three non-linear models namely, Gompertz, Bertalanffy and Logistic models were evaluated. Goodness of fit for all the models were checked using coefficient of determination (R2), adjusted coefficient of determination (Adj-R2), mean square error (MSE), mean absolute error (MAE) and Akaike information criterion (AIC). The Bertalanffy model most accurately characterized the growth trend in males, females and pooled sex data. The study revealed that this model may be used to ascertain the average body weights in Aseel chicken under random mating. The investigation has generated baseline data on growth modelling of random bred groups and may be used in similar investigations on other native chicken breeds. Keywords: Aseel, Bertalanffy model, Gompertz model, Growth models, Logistic model.
View
... Besides more egg production, optimum growth in body weights is also an important attribute to the farmers for promoting RIR in rural livelihood. It is necessary to have knowledge of factors influencing its growth in body weights, an important factor contributing to the profitability in poultry production (Kausar et al. 2016). It is also desirable to have estimates of genetic and phenotypic parameters afresh in each generation for each population, because the estimates vary from one population to another and at different times (Barot et al. 2008). ...
Genetics of body weights in a control line of Rhode Island Red grower chicken
Article
Full-text available
  • Aug 2017
Investigating a control line population of RIR grower chicken to estimate genetic and non-genetic parameters of its body weights, it was summarized that body weight growth is influenced by its sire-inheritance and sex where males being heavier throughout the ages. Significant hatchdifferences and chick weight-regression effect were also noticed up to the age of 16th week. The heritability, genetic and phenotypic correlation estimates were mostly with moderate to high magnitudes. Thus the present findings could serve as the pre-requisites for chalking out of breeding strategies for genetic improvement of the chicken genotype.
View
... Growth curve describes the growth pattern over time by non-linear mathematical functions which convert weight-time information into certain number of growth parameters. These parameters provide biological interpretations, which can be used in deciding the nutritional practices, age for slaughtering and appropriate revisions in selection strategies (Bathaei and Leroy, 1996; Lambe et al., 2006;Kausar et al., 2016). It can be useful for identifying better animals in advance in order to keep superior animals at farm (Malhado et al., 2009;Canaza-Cayo et al., 2015). ...
Genetic Parameter Estimates for Growth Curve Characteristics of Deccani Sheep
Article
Full-text available
  • Jan 2017
Deccani sheep is an important breed of sheep found in semi-arid zones of India. The present study was based on the monthly body weights of 753 Deccani lambs maintained at ICAR-Network project during 2011 to 2015. The objectives of the study were to find the effect of various factors on the growth curve parameters estimated by Brody model and to estimate the genetic parameters of the growth curve. The Brody function was used to estimate the growth curve parameters for each lamb separately and then, growth parameter estimates were tested if there was significant influence of fixed effects of various factors. The genetic parameters of growth curve parameters were estimated under restricted maximum restricted likelihood procedure. The least-square estimates for mature weight (A), proportion of mature weight attained after birth (B) and maturation rate (k) were 30.68 ± 0.26, 1.82 ± 0.04 and 0.66 ± 0.01, respectively. Year of birth, season of birth and sex of lamb had significant influence on all growth curve parameters. The heritability estimates for A, B and k were low as 0.08 ± 0.06, 0.11 ± 0.07 and 0.07 ± 0.06, respectively. The phenotypic and genetic correlation between A and k was negative which indicated that early maturing leads to lower mature weights.
View
COMPARISON OF NONLINEAR MODELS FOR DESCRIBING THE GROWTH OF LOCAL CHICKENS RAISED IN TOGO
Article
Full-text available
  • Oct 2024
Growth is a trait of economic importance in chicken genetic improvement that is often described by mathematical models having biologically significant parameters. The local chicken,commonly raised in rural areas of Togo, is generally described as a slow-growing chicken, but little is known about his growth parameters. Therefore, this study was designed to determine the nonlinear model that best describes the growth curve of local chickens. The Gompertz, Logistic and Richards growth models were used to fit growth weights recorded from 126 males and 150 female chickens using GCFIT program of SIMFIT package (version 8.1.4).Coefficient of determination (R2), mean square error (MSE), Akaikes information criterion (AIC) and Bayesian information criterion (BIC) were used as criteria for models comparison. By estimating the highest value of R² and the lowest values of MSE, AIC and BIC, Gompertz model appeared to be the best-fitting model for local chickensgrowth. The estimated asymptotic weight was 1992 g for males and 1367 g for females. Thus, male and female chickens weighed 732.86g and 502.98g and aged 10.57 weeks and 9.77 weeks respectively at inflection point. Maturation rates of the chicks were similar (0.139 week-1 for males and 0.141 week-1for females). The growth parameters of this model may therefore be useful in defining selection criteria for local chickens in Togo.
View
Microsatellite polymorphism, immunocompetence profile and performance evaluation of Rhode Island Red chicken and its crosses.
Thesis
Full-text available
  • Apr 2013
This investigation envisaged was carried out to study microsatellite polymorphism in two Rhode Island Red (RIR) strains, viz. selected (RIRS) and control (RIRC) strains, immunocompetence profiles and growth and production performances of RIR strains, and their crosses and also estimate association of microsatellite alleles and immunocompetence traits with various economic traits in RIR strains and strain-crosses. The Microsatellite analysis revealed 2 - 7 alleles in RIRS and 2 - 9 alleles in RIRC strains at 24 microsatellite (MS) loci with their molecular sizes of ranging from 84 bp at MCW0051 in RIRC to 276 bp at MCW0050 MS loci in RIRC, with varied frequencies. The mean observed number of alleles (Ha) per locus was 4.04 ± 0.23 in RIRS and 4.42 ± 0.33 in RIRC. A total of 97 alleles in RIRS and 106 in RIRC were resolved in 24 loci. Alleles’ frequency ranged from 0.083 to 0.667 in RIRS and 0.083 to 0.833 in RIRC. The most frequent allele was 116 bp allele (66.7%) at MCW0051 locus in RIRS and 173 bp allele (88.3%) at MCW0059 locus in RIRC. Analysis revealed 34.02% alleles with frequency range of 0.083 to 0.667 in RIRS and 39.62% alleles with a frequency range of 0.083 to 0.883 in RIRC as population specific alleles. Loci ADL0136 and MCW0004 revealed highest number of specific alleles in RIRS (5) and RIRC (6) strains. Average heterozygosity (Nei’s H) per locus ranged from 0.4800 (MCW0059) to 0.8056 (ADL0267) with a mean of 0.6753±0.019 in RIRS and from 0.2778 (MCW0059) to 0.8750 (ADL0136) with a mean of 0.6776±0.028 in RIRC. All the loci were polymorphic and PIC ranging from 0.2392 (MCW0059) to 0.8620 (ADL0136). The effective number of alleles (Ne) and Shannon’s information index (I) averaged 3.3230±0.190 and 1.2455 ± 0.057 in RIRS and 3.6616±0.315 and 1.2992 ± 0.080 in RIRC strains. The mean effective number of alleles (Ne) maintained in both the populations was less than the mean observed number of alleles (Na) in both the populations, which indicated that allele frequencies were widely distributed. Na, Ne, I statistics and mean observed heterozygosity (Ho) of 0.6306±0.3901 in RIRS and 0.6528±0.4345 in RIRC indicated that RIRC was more diverse population than the RIRS strain. The expected heterozygosity (He) ranged from 0.5053 (MCW0059) to 0.8421 (MCW0004) in RIRS and from 0.2899 (MCW0059) to 0.9130 (ADL0136) in RIRC. Mean Ho < mean He indicated that RIRS and RIRC populations were not in Hardy-Weinberg equilibrium but under the influences of some forces like selection for some traits associated with microsatellite loci. Write’s F-statistics analysis revealed that 62.50% loci in RIRS and 58.33% loci in RIRC contributed moderate to high negative FIS index value indicating heterozygote excess for those loci. Mean FIS was indicative of heterozygosity deficiency and FIS of 0.1077 on pooled populations indicated overall 10.77% heterozygosity deficit. A mean FIT of 0.1719 indicated 17.19% inbreeding co-efficient favoring homozygosity over the two populations whereas 54.14% loci exhibited fairly moderate negative FIT estimates indicative of excess in heterozygosity in those loci over the populations. FST statistic indicated that 10.18% of microsatellite variation between the two populations. The mean FIS, FIT and FST for all the microsatellites together in the overall population was 0.1077, 0.1719 and 0.1018, respectively. The original measures (Nei, 1972) of genetic identity and genetic distance were 0.5264 and 0.6418. The dendrogram based on Nei’s (1972 &1978) genetic distance showed the genetic diversity between two populations with moderate distance reflecting 29.64 - 32.09% common inheritance. In the immunocompetence (IC) profiles, the least squares (LS) means of HA titre were 8.854 ± 0.359, 10.379 ± 0.616, 6.719 ± 0.548, 6.001 ± 0.441 and 5.739 ± 0.436 in RIRS, RIRC, RIRW, CARI-Sonali (HR) and CARIDebendra (CD), respectively. Corresponding estimates of serum lysozyme concentration were 6.503 ± 0.591, 5.160 ± 0.357, 6.730 ± 0.693, 5.692 ± 0.324 and 6.031 ± 0.213 μg/ml, and of serum IgG concentration were 6.663 ± 0.455, 7.761 ± 0.380, 7.622 ± 0.391, 5.167 ± 0.413 and 5.999 ± 0.347 μg/μl. Sire effect was significant (P≤0.05) on serum lysozyme and IgG concentrations in RIRS only. The sex did not have significant (P>0.05) effect on any IC traits in any genotype. All the three IC traits differed significantly (P≤0.05) in five genotypes. The trend of LS means of HA titre was RIRC > RIRS > RIRW > HR > CD. For serum lysozyme concentration, it was RIRW > RIRS > RIRC > CD > HR and for IgG, it was RIRC > RIRW > RIRS > CD > HR. Pure strains demonstrated better immune-competence than strain- crosses. The heritability (h2) estimates of HA titre, serum lysozyme and serum IgG concentrations were of low to moderate in magnitude. The estimates of genetic correlations were less precise due to high standard errors. Phenotypic (rP) correlations among various IC traits in pure strains and strain-crosses were low in magnitude and showed with no definite trend. Percent fertility pooled over hatches was 75.86, 79.03, 79.34, 70.83 and 60.64% of the parents of RIRS, RIRC, RIRW, HR and CD chicks, respectively. RIRW demonstrated highest percent fertility followed by RIRC > RIRS > HR > CD. Percent hatchabilities, calculated on the basis of total egg set (TES), pooled over hatches, were 57.46, 68.51, 67.80, 55.58 and 53.09%, respectively for RIRS, RIRC, RIRW, HR and CD chicks. Percent hatchability on the basis of fertile egg set (FES), pooled over hatches, were calculated as 75.65, 86.62, 85.27, 78.44 and 87.54 %, respectively for RIRS, RIRC, RIRW, HR and CD chicks. Sire, genotype, sire within genotype (nested) and hatch effects on all the growth traits were mostly significant (P≤0.05). Genotype-sex interaction effect was significant (P≤0.05) on all the growth traits from 8th week onwards. Significant (P≤0.05) sire effect (for all the layer production traits in RIRS, not in RIRC strain), sire within genotype effect (excepting EW28), hatch effect (excepting AFE in CARI-Debendra, EWs in all genotypes, EP40 in RIRC and CARI-Debendra) and genotype effect were reflected on all the layer production traits. Chick weight as regression had significant (P≤0.05) effect on growth traits at 1st (pure strains) or 2nd (crosses) weeks onwards up to 20th week in RIRS, 16th week in RIRC, 12th week RIRW, 20th week in CARI-Sonali, and 6th week in CARI-Debendra genotype. Housing body weight regression effect was significant on AFE and other layer production traits up to 40th week excepting EW28 & EP40 in RIRC, EP in CARI-Sonali, EWs & EP40 in CARI-Debendra. Chick weight was highly heritable. Various growth traits were found to have heritabilities from low to moderate in magnitude. Various layer production traits were found to have heritability from low to moderate in magnitude. Genetic correlations among various growth traits were moderate to high in magnitude and mostly positive. Genetic correlations (rG) among layer production traits estimated moderate to high magnitude coefficients but variable direction in both pure strains. AFE remained positive rG with EP40 invariably in both pure strains but could not follow any specific trend with EWs which remained negative rG with EP40 in RIRS but positive in RIRC. Phenotypic correlations (rP) among growth traits were positive and low to high in magnitude. The rP among layer production traits were either positive or negative but low to high in magnitude. AFE remained negative rP with EP40 invariably in all genotypes but could not follow any specific trend with EWs which had negative rP with EP40 in RIRS and CARI-Sonali but positive in RIRC, RIRW and CARI-Debendra. Some of the microsatellite genotypes had significant (P≤0.05) effect on layer economic traits. Effect of IC traits on layer production traits revealed significant (P≤0.05) effect. High HA titre level had significantly (P≤0.05) greater LS means of EW40 than medium or low levels of HA titre in RIRS. Likewise, high/medium serum IgG level had significantly (P≤0.05) greater LS means of EW28 than low level of serum IgG, and high serum IgG level had significantly (P≤0.05) greater LS means of EW40 than medium or low levels of serum IgG in RIRS. In RIRC, high serum IgG level had significantly (P≤0.05) greater LS means of BW20 than medium level of serum IgG. In HR, high/medium HA titre levels had significantly (P≤0.05) greater LS means of EW20 than low level of HA titre, and medium serum lysozyme level had significantly (P≤0.05) lower LS means of AFE than high level of serum lysozyme. In CD, medium serum lysozyme level had significantly (P≤0.05) greater LS means of BW40 than high level of serum lysozyme, and high serum IgG level had significantly (P ≤ 0.05) greater LS means of EW28 than medium or low levels of serum IgG. This investigation concluded that the isolated strain- specific alleles of microsatellite loci could be used for strain-differentiation. RIRC was more diverse than RIRS based on Nei’s H and Mean Na, Ne & I statistics. However, both populations demonstrated that they were not in H-W equilibrium but under the influences of some forces like selection. Most of the MS loci were moderate to highly informative in both strains excepting MCW0059 in RIRC (PIC<0.25). Both strains also showed high genetic identity and genetic distance which might be due to their similar genetic base. Trend of LS means of HA titre was RIRC > RIRS > RIRW > CARI-Sonali > CARI-Debendra; serum lysozyme concentration had RIRW > RIRS > RIRC > CARI-Debendra > CARI-Sonali and serum IgG showed RIRC > RIRW > RIRS > CARI-Debendra > CARI-Sonali. Pure strains demonstrated better immunocompetence than crosses. CARI-Debendra demonstrated higher growth traits than CARI-Sonali > RIRS > RIRW > RIRC. CARI-Sonali pullets demonstrated (P≤0.05) least AFE than RIRS < RIRW < CARI-Debendra ≤ RIRC. CARI-Debendra ≥ CARI-Sonali pullets had (P≤0.05) higher EW28 & EW40 than RIRS > RIRC and RIRS ≥ RIRW and RIRW ≥ RIRC. CARI-Sonali pullets had (P≤0.05) higher EP40 than RIRS > RIRW ≥ CARIDebendra > RIRC. Effect of chick weight began at 1st week in pure strains or 2nd week in crosses and continued up to 20th week in RIRS, 16th week in RIRC, 12th week in RIRW, 20th week in CARI-Sonali and 6th week in CARI-Debendra. Thus illustrating high egg producing layer birds (CARI-Sonali > RIRS for EP40) had inconsistent CW regression effect on its growth traits as compared to low egg producing birds. Effect of housing body weight (BW20) was significant on AFE and continued up to 40th wk on all layer production traits excepting a few traits. Heritability (h2) estimates were moderate to high in CW & growth traits, and moderate in layer production traits across all genotypes. Genetic (rG) and phenotypic (rP) correlations were low to moderate in growth and production (GP) traits across all genotypes. Significant (P≤0.05) associations between microsatellites and production traits have been observed for some loci indicating phenomenon of linkage (LD) between them and for conformation requires investigation on large numbers of individuals. Significant (P≤0.05) associations between IC levels and production traits have been observed.
View
Non-linear models for growth curves in Large White turkeys
Article
Full-text available
  • Jan 2005
The present work aimed to model the growth curves of male and female turkeys with respect to their live weight-age relationships and to determine a non-linear model explaining their growth curve better. For this purpose four different non-linear models were used to define growth curves of turkeys, namely Gompertz, Logistic, Morgan-Mercer-Flodin (MMF), and Richards. The coefficients of determination for these models were 0.9975, 0.9937, 0.9993. and 0.9966 for females and 0.9974, 0.9933, 0.9993, and 0.9969 for males, respectively. Considering model selection criteria, Gompertz, Logistic and Richards models seen to be suitable models for explaining Large White turkey growth.
View
An evaluation of different growth functions for describing the profile of live weight with time (age) in meat and egg strains of chicken
Article
Full-text available
  • Nov 2003
A total of 86 profiles from meat and egg strains of chickens (male and female) were used in this study. Different flexible growth functions were evaluated with regard to their ability to describe the relationship between live weight and age and were compared with the Gompertz and logistic equations, which have a fixed point of inflection. Six growth functions were used: Gompertz, logistic, Lopez, Richards, France, and von Bertalanffy. A comparative analysis was carried out based on model behavior and statistical performance. The results of this study confirmed the initial concern about the limitation of a fixed point of inflection, such as in the Gompertz equation. Therefore, consideration of flexible growth functions as an alternatives to the simpler equations (with a fixed point of inflection) for describing the relationship between live weight and age are recommended for the following reasons: they are easy to fit, they very often give a closer fit to data points because of their flexibility and therefore a smaller RSS value, than the simpler models, and they encompasses simpler models for the addition of an extra parameter, which is especially important when the behavior of a particular data set is not defined previously.
View
Nonlinear stochastic models for describing growth of soybean production in Madhya Pradesh and India
Article
  • Dec 2014
The aim of this study is to form an analysis of nonlinear stochastic models in soybean production in Madhya Pradesh and India. The data set used for this analysis was taken from soybean processor association (SOPA) Indore. Eight nonlinear stochastic models were taken: Gompertz, Logistic, Richard, Janoschek, Champman, Morgan-Mercer-Flodin (MMF), Boltzmann and Weibull. For model fitting performance, we used four comparison criteria; coefficient of determination (R2), sum of Squares of error (SSE), root mean squares error (RMSE) and mean relative error (MRE). The results indicated that Weibull and MMF model are more useful than other non-linear models to estimate soybean production in Madhya Pradesh and also in India.
View
View
Nonlinear models to describe growth pattern of Indian native chickens
Article
  • Jun 2008
Aseel is a famous bird of Indian native chicken and is well known for its high stamina, majestic gait and dogged fighting qualities. In addition, it has plenty delicious flavored meat due to its biggest body size among the Indian native breeds of chickens accompanied by broader breast. About 13 varieties of Aseel are found in the country out of which Kagar and Peela are being maintained at the Central Avian Research Institute, Izatnagar. The average body weight of Aseel Peela at 0, 8, 20, 30 and 40 weeks of age for male and female chickens were 36.18±0.56 g, 35.89±0.53 g; 402.00±12.84 g, 385.37+11.20 g; 1550.38+34.72 g, 1479.63±43.66 g; 2246.25±57.54 g, 2097.96±71.51 g and 2627.62±71.94 g and 2449.07±103.72 g respectively. The male birds grew faster than female birds and both attained the maximum weight gain at 15th week of age. Seven nonlinear functions were fitted to average weekly body weights of male and female birds. Gompertz function was found to be the best function to describe the growth pattern of these chickens.
View
Nonlinear Growth Models for Body Growth of Vanaraja Poultry Birds
Article
  • Jul 2012
The present study aims to find out a suitable growth model of Vanaraja, a new bird of poultry. Five different nonlinear growth models were attempted, among which Richards growth model was seemed to be the best model to represent the growth pattern of the poultry bird. The asymptotic weight of the bird given by Richards model was 2.516 kg approximately.
View
Nonlinear growth curve analysis: Estimating the population parameters
Article
  • Mar 1986
A model for analysing longitudinal growth data with covariates is proposed. It assumes that the data for each individual follow a nonlinear growth model, with parameters unique to that individual. The parameters of individuals are assumed to vary in the population, with the population mean of each parameter dependent upon covariates. Several simple, two-stage methods of estimation are described and compared for making inferences about the effects of covariates upon growth. We also present a method of model validation for nonlinear models. These two-stage methods solve the inference problem we encounter when performing multiple cross-sectional analyses of the effects of covariates along the age scale. In addition, these two-stage methods permit us to study the effects of covariates upon growth rate, or other, possibly nonlinear, growth parameters. We find that some methods are better than others at reproducing the covariate effects observed in a series of age-specific analyses of the data.
View
View
An adjustment model of Logistic form to describe the growth pattern of chickens
  • Prasad Shiv
  • D P Singh
Prasad Shiv and Singh D P. 2006.An adjustment model of Logistic form to describe the growth pattern of chickens. Indian Journal of Poultry Science 41 (3): 280-82.