Evaluation of Growth Parameters of Broiler Chickens Fed with Different Dietary Levels of Hemp Seed (Cannabis sativa L.) using Gompertz Model Compare with Artificial Neural Network

This study was conducted to evaluate the effects of different dietary levels of Hemp seed (HS) (Cannabis sativa-L.) on growth parameters of broiler chickens was estimated using Gompertz and Artificial neural network models.

In this study, 192 male broiler chicks (1 d old-Ross 308) were randomly assigned to a completely randomized design with 4 dietary treatments: control (without HS), 2.5, 5 and 7.5% HS in 4 replications (12 birds/pen). The chickens had freely accessed to drinking water and fed ad-libitum. To estimate growth parameters, cumulative body weight of birds was fitted to Gompertz model.

The results showed that different dietary levels of Hemp seed had no significant effect on growth parameters of broiler chickens (p>0.05). But growth rate at first to fifth weeks affected by different dietary levels of Hemp Seed (p<0.05). Results of comparison of models showed that non-linear Gompertz model had higher R2, and lower MSE, MAD, MAPE and bias compared with artificial neural network, that had better estimate of weight of broiler chickens.

The results of this study showed that different levels of Hemp seed diets had a significant effect on growth parameters of broilers including relative growth rate, live weight at maturity, turning point of growth curve and body weight at turning point of growth curve. Growth rate in the first to fifth weeks of broiler breeding period was significantly affected by the addition of hemp seed diets. On the other hand, the results of this study showed that the Gampertz model was able to estimate the 42-day-old weight of broilers more accurately than the artificial neural network model.