This paper applies Auto-Regressive Integrated Moving Average model (ARIMA) with (multivariate model) and without (univariate model) explanatory variables to predict the fish catch in Lake Manzala , Egypt . Bootstrap technique has been applied to allow one to judge the uncertainty of estimators obtained from the suggested ARIMA model, without prior assumptions about the underlying probability distributions. This method is based on generating 1000 random samples with replacement from the original observations. Then refit the suggested ARIMA model to each sample to end with a probability distribution of the model parameters, which can be used to estimate the bias of the parameters. Also, Jackknife technique has been applied after bootstrapping the ARIMA model. Jackknife-after-bootstrap (JAB) technique aimed to estimate the bias of the bootstrap estimates, by deleting each observation in turn to obtain nestimates based on  observations; this procedure has been repeated for each bootstrap sample. The main aim of applying these techniques is to try to minimize the bias of the estimation of ARIMA model, so we can get accurate estimation of the parameters which consider the most accurate tool to help the decision maker to apply proper management policies to optimize the fishing effort and protect the fish stock. Evidence from the Egyptian fisheries revealed that using ARIMA coupled with both bootstrap and JAB leads to the most accurate prediction in the perspective especially with time series data.

auto-regressive integrated moving average (ARIMA), bootstrap, Jackknife-after-bootstrap, multivariate model, univariate model, periodogram, regressor variables.