Short-term forecasting of halibut CPUE: Linear and non-linear univariate approaches

In the present paper, two univariate forecasting techniques were tested to evaluate the short-term CPUE capacity forecast for Pacific halibut, Hippoglossus stenolepis (Pleuronectidae). The first methodology, based on the Box–Jenkins approach (autoregressive integrated moving average models [ARIMA models]), assumes a linear relationship between the time series data. The second methodology, using artificial neural network models (ANNs), enables highly non-linear processes to be modelled. The best results from a seasonal ARIMA model indicated that one non-seasonal autoregressive term combined with a non-seasonal moving average term and a seasonal moving average term was suitable to explain a variance level of 32.6% in the validation phase, providing statistically acceptable but insufficiently satisfactory estimations. The configuration of the best ANN model (three autoregressive terms in the input layer and five neurons in the hidden layer) provided a significant improvement in the independent validation phase (91% of the variation explained), indicating a clear non-linear relationship between variables. Modelling of the abundance indices is a useful tool for understanding the dynamics of populations and may enable short-term quantitative recommendations for fisheries management to be made.     

Performance of a closed recirculating system with foam separation, nitrification and denitrification units for intensive culture of eel: towards zero emission

The development of a closed recirculating aquaculture system that does not discharge effluents would reduce a large amount of pollutant load on aquatic bodies. In this study, eel were reared in a closed recirculating system, which consisted of a rearing tank, a foam separation unit, a nitrification unit and a denitrification unit. The foam separation unit has an inhalation-type aerator and supplies air bubbles to the rearing water. The growth of eel, which were fed a commercial diet, was satisfactory, with gross weight increases of up three times in 3 months. The survival rate under the congested experimental conditions was 91%. The foam separation unit maintained oxygen saturation in the rearing water at about 80%. Furthermore, fine colloidal substances were absorbed on the stable foam formed from eel mucus and were removed from the rearing water by foam separation. Ammonia oxidation and the removal of suspended solids were accomplished rapidly and simultaneously in the nitrification unit. The ammonia concentration and turbidity were kept at less than 1.2 mg of N per litre and 2.5 units, respectively. When the denitrification process was operated, nitrate that accumulated in the rearing water (151 mg of N per litre) was reduced to 40 mg of N per litre. The sludge was easily recovered from the nitrification and denitrification tanks, and the components were found suitable as compost. Based on these results, the intensive aquaculture of freshwater fish such as eel can be achieved using a closed recirculating system without emission.     

Modeling and forecasting pelagic fish production using univariate and multivariate ARIMA models

ABSTRACT:   Univariate and multivariate autoregressive integrated moving average (ARIMA) models were used to model and forecast the monthly pelagic production of fish species in the Mediterranean Sea during 1990–2005. Autocorrelation (AC) and partial autocorrelation (PAC) functions were estimated, which led to the identification and construction of seasonal ARIMA models, suitable in explaining the time series and forecasting the future catch per unit of effort (CPUE) values. Univariate and multivariate ARIMA models satisfactorily predicted the total pelagic fish production and the production of anchovy, sardine, and horse mackerel. The univariate ARIMA models demonstrated a good performance in terms of explained variability and predicting power. The current findings revealed a strong autoregressive character providing relatively high R ² and satisfactory forecasts that were close to the recorded CPUE values. The present results also indicated that the multivariate ARIMA outperformed the univariate ARIMA models in terms of fitting accuracy. The opposite was evidenced when testing the forecasting accuracy of the two methods, where the univariate ARIMA models overall performed better than the multivariate models. The observed seasonal pattern in the monthly production series was attributed to the intrinsic nature of the pelagic fishery. As anchovy, sardine, and horse mackerel represent main target species in the Mediterranean pelagic fishery, the findings of the present study provided direct support for the potential use of accurate forecasts in decision making and fisheries management in the Mediterranean Sea.     

Using a macroalgae Ulva pertusa biofilter in a recirculating system for production of juvenile sea cucumber Apostichopus japonicus

A simple indoor recirculating system for production of juvenile sea cucumber (Apostichopus japonicus) was operated on a commercial scale for 90 days during winter. The system consists of three 70 m³ sea cucumber rearing tanks and one biofilter tank where macroalgae (Ulva pertusa) was used as a biofilter in order to reduce water requirements. Effluent from the sea cucumber tanks drained into the macroalgae biofilter tank and were then returned to the sea cucumber tanks by a discontinuous-flow recirculation system. Survival and growth rates in the sea cucumber culture tanks were similar to those in the control tank (with one water exchange per day). The survival rate averaged about 87%. The average body weight increased from 3.5 ± 0.3 g to 8.1 ± 0.8 g and total sea cucumber biomass production over the experimental period was 745 g m⁻² after initial stocking densities of 375 g m⁻². The growth rate of U. pertusa was 3.3% day⁻¹. U. pertusa was efficient in removing toxic ammonia and in maintaining the water quality within acceptable levels for sea cucumber culture; there were only small daily variations of temperature, pH and DO. The U. pertusa tank removed 68% of the TAN (total ammonia-nitrogen) and 26% of the orthophosphate from the sea cucumber culture effluent; the macroalgae biofilter removed ammonia at an average rate of 0.459 g N m⁻² day⁻¹. It would be efficient to use the U. pertusa biofilter in a recirculating system for production of A. japonicus juveniles in winter.     

An experimental study on nitrification biofilm performances using a series reactor system

Total ammonia nitrogen (TAN) concentration is often a key limiting water quality parameter in intensive aquaculture systems. Removing ammonia through biological filtration is thus the first objective in recirculating aquaculture system design. In this study, the performance characteristics of a steady-state nitrification biofilm were explored using a series of reactors. Four nitrification kinetics parameters were estimated using the data collected from the experimental system, including minimum TAN concentration, half saturation constant, maximum TAN removal rate and maximum specific bacterial growth rate. Experimental data showed that a minimum TAN concentration was needed to support a steady-state nitrification biofilm. For the temperature of 27.2°C, the mean minimum TAN concentration was 0.07 mg/l. For a single substrate-limiting factor, the relationship between TAN removal rate (R) and TAN concentration (S) was represented by an empirical equation [R=1859(S−0.07)/(S+1.93)]. The characteristics of nitrite oxidation were also demonstrated by the experiment system. The results of this study will help to better understand the characteristics of nitrification biofilters applied in recirculating aquaculture systems.     

Prediction of shrimp growth using an artificial neural network and regression models

Uncertainty and lack of information about the future make it difficult for shrimp farmers to develop and plan harvesting schedules. To do this effectively, farmers should be able to predict shrimp growth. A reliable prediction of growth and survival would also give farmers a better insight into future productivity and profitability. Linear and nonlinear regression models have been used to estimate growth of different types of animals. These models include theoretic guesses and hypotheses about the underlying laws that govern the system from which data are generated. Compared to such models, artificial neural networks (ANNs) make a few priori assumptions about the models and suited for predicting animal growth. This study evaluated the potential of an ANN as an alternative to regressions models for predicting shrimp growth. Empirical data were collected from 9 commercial shrimp farms in the Bushehr Province of Iran. The results showed that the ANN performed better compared to linear and nonlinear regression models for predicting the growth of farmed shrimp.     

Research of dissolved oxygen prediction in recirculating aquaculture systems based on deep belief network

Recirculating aquaculture has received more and more attention because of its high efficiency of treatment and recycling of aquaculture wastewater. The content of dissolved oxygen is an important indicator of control in recirculating aquaculture, its content and dynamic changes have great impact on the healthy growth of fish. However, changes of dissolved oxygen content are affected by many factors, and there is an obvious time lag between control regulation and effects of dissolved oxygen. To ensure the aquaculture production safety, it is necessary to predict the dissolved oxygen content in advance. The prediction model based on deep belief network has been proposed in this paper to realize the dissolved oxygen content prediction. A variational mode decomposition (VMD) data processing method has been adopted to evaluate the original data space, it takes the data which has been decomposed by the VMD as the input of deep belief network (DBN) to realize the prediction. The VMD method can effectively separate and denoise the raw data, highlight the relations among data features, and effectively improve the quality of the neural network input. The proposed model can quickly and accurately predict the dissolved oxygen content in time series, and the prediction performance meets the needs of actual production. When compared with bagging, AdaBoost, decision tree and convolutional neural network, the VMD-DBN model produces higher prediction accuracy and stability.     

Process modelling of an intensive aquaculture system

The application of process modelling concepts to an intensive recirculating aquaculture system is described. Advanced Continuous Simulation Language (ACSL) for describing the unsteady behaviour of complex continuous flow system is applied to a simplified recycle fish rearing system comprising a rearing tank and a treatment unit. Use of the technique was validated by comparison with analytical solutions of the unsteady state mass balance equations for the transfer and removal of ammonia. The language is successfully implemented to study the dynamic behaviour of ammonia concentration and in response to changes imposed upon the system. The effects of choice of design and operating parameters upon the response of the system are illustrated. Parameters studied include vessel size, lag, flowrate, recycle ratio and mixing regime.     

Cyclical fluctuations in ammonia and nitrite-nitrogen resulting from the feeding of turbot, scophthalmus maximus (L.), in recirculating systems

Daily and weekly cyclical fluctuations in ammonia (NH₄---N μg litre⁻¹) and nitrite---nitrogen (NO₂---N μg litre⁻¹) were shown to occur in marine recirculating systems containing juvenile turbot, Scophthalmus maximus (L.). A stepwise multivariant regression analysis, computed for the prefeeding morning base levels, showed that ammonia---nitrogen concentration in the fish tanks was related to inflow ammonia---nitrogen concentration, water flow rate, water temperature, mean fish weight and fish biomass. Influent ammonia concentration and water flow rate were by far the most important variables accounting for 68·0% and 16·7% of the variation in the regression respectively.

The fluctuations, which occurred at concentrations thought to be tolerated by the fish, were attributed to the amount and timing of feeding. The significance of the existence of such fluctuations was discussed together with the importance of their amplitude and frequency.     

Is it possible to differentiate between environmental and fishery effects on abundance‐biomass variation? A case study of blackspot seabream (Pagellus bogaraveo) in the Strait of Gibraltar

We assessed the potential for simulation and modelling of the blackspot seabream (Pagellus bogaraveo) population in the Strait of Gibraltar to discriminate the environmental effects of fishery impacts. A discrete biomass–abundance dynamic model was implemented to obtain a simulated monthly time series of blackspot seabream biomass. On this simulated time series, autoregressive integrated moving average (ARIMA) models were fitted. The best ARIMA fit provided a significant correlation of 0.76 and persistence index higher than 0.85. The proportion of variance non‐explained by the ARIMA models was correlated with a time series of sea surface temperature (SST) and North Atlantic Oscillation (NAO). The analysis of global, annual and winter correlation between the proportion of variance not explained by the ARIMA models and environmental variables showed that significant associations were not detected over the full time series. Our analysis therefore suggests that overexploitation is the main factor responsible for the commercial depletion of blackspot seabream in the Strait of Gibraltar.     

Spatial organisation of European eel (Anguilla anguilla L.) in a small catchment

Abstract –  Modelling-governing patterns of European eel ( Anguilla anguilla L.) distribution of four eel size classes (<150, 150–300, 300–450 and >450 mm) in the Frémur basin (northwest France) was done using artificial neural network (ANN) techniques and ecological profiles. Our results demonstrate the high predictive power of the ANN models. Some macro- and microscale factors, such as distance from the sea, depth and flow velocity, have the most significant influence on the models. Influence of distance from the sea appears to be very different from the spatial organisation usually described in river systems. In fact, the general tendencies of total eel densities according to the distance from the sea showed that densities increase weakly upstream. Another outcome was the variations in habitat preference according to the eel size, even if this species is spread over practically every type of microhabitat. Small eels were mainly found in shallow habitats with strong abundance of aquatic vegetation, whereas large eels tend to be found in intermediate to high depth with small to intermediate abundance of aquatic vegetation. Finally, we hypothesise that European eels change behaviour and microhabitat characteristic preference around a size of 300 mm.     

Productive performance of juvenile freshwater prawns Macrobrachium rosenbergii in biofloc system

The aim of this study was to compare two rearing systems for freshwater prawn Macrobrachium rosenbergii: one with use of a recirculating aquaculture system with biofilters (RAS) and another with use of microbial flocs (F). Thirty postlarvae of freshwater prawn with an initial average weight of 0.13 ± 0.05 g were randomly stocked in six experimental units with 0.20 m² and volume of 50 L. The experiment lasted thirty days. Dissolved oxygen, temperature and pH were monitored daily; ammonia concentration was determined three times per week; nitrite concentration, alkalinity and hardness were measured weekly. For the formation of microbial floc, molasses was used to keep the ammonia concentrations within safe levels for prawn farming. The variables of water quality remained within the suitable range for the production of the species, except for ammonia concentrations at the F treatment, which exceeded the safe levels. At the end of the experiment, the following parameters were evaluated: survival, specific growth rate, weight gain and feed conversion rate. Differences were found only in feed conversion rate with better values on RAS treatment. The microorganisms present in the RAS and F treatment were also evaluated. The densities of rotifers, amoebas and total bacteria were higher at the F treatment although the same organisms were found at the RAS treatment. The results of this study showed the possibility of rearing M. rosenbergii in biofloc system technology.     

Ammonia reduction in seawater by Yucca schidigera extract: efficacy analysis and empirical modelling

Yucca schidigera is a plant native to southwestern USA and Mexico. Its extract has been used in the livestock industry to control ammonia accumulation in animal holding facilities, and to reduce ammonia concentration in animal excreta. This study investigated the potential and effectiveness of Y. schidigera extract (YUPE) for ammonia reduction in seawater. A dose–response experiment was conducted to determine the effect of different concentrations of YUPE at 0, 18, 36, 72 and 108 mg L⁻¹ on total ammonia nitrogen (TAN) at 1, 3 and 9 mg L⁻¹. At a higher YUPE dosage rate, higher TAN reduction was observed, and TAN reduction was highest during the first 12 h, and decreased thereafter. A stepwise multiple linear regression that included the initial TAN, YUPE concentration and time was developed, which accurately predicted empirical TAN concentrations. Applications of this model for ammonia management strategies were formulated for hypothetical tiger prawn (Penaeus monodon) rearing conditions. YUPE's efficacy for ammonia reduction, natural origin and safety make YUPE a potentially suitable compound for water quality management in mariculture.     

Modeling inflow rates for the water exchange management in semi-intensive aquaculture ponds

Several linear and non-linear models for centralized remote-control systems that can support decision making of semi-intensive aquaculturists concerning the inflow rates to the ponds were evaluated. These models were: multiple linear regressions (MLRs), generalized additive models (GAMs), artificial neural networks (ANNs) and fuzzy logic controllers (FLCs). These modeling techniques were applied in a semi-intensive gilthead seabream (Sparus aurata) fishfarm located in southern Spain. The water temperature, ammonia concentration, turbidity and dissolved oxygen concentration in the ponds were measured and used as independent variables. Of all the approaches employed to simulate the actual water exchange operation in the ponds, the best fits were obtained using ANN and FLC models with only three input variables (turbidity measured at the input of the ponds and dissolved oxygen measured at the input and output of the ponds). These models provided levels of correlation between 0.73 and 0.75. In contrast, the best GAM and MLR models provided correlation coefficients of only 0.38 and 0.33, respectively. In spite of the results being statistically significant, the explained variance levels obtained indicate how difficult it is to capture the experience and knowledge of the aquaculturist concerning the operation of the water exchange in the ponds for maintaining the water quality in these production systems.     

Growth performance of disk abalone Haliotis discus hannai in pilot- and commercial-scale recirculating aquaculture systems

The study investigated the growth performance of abalone from juvenile to marketable size in a commercial-scale recirculating aquaculture system. The rearing system consisted of 12 raceways (4.0 × 0.8 × 0.6 m) with a protein skimmer and a submerged biofilter for juveniles and 10 raceways (6.6 × 1.3 × 0.6 m) with a protein skimmer and a trickling biofilter for on-growing. Sea mustard (Undaria pinnatifida) and kelp (Laminaria japonica) were fed to the abalone. The total weight of abalone in the recirculating aquaculture system at the juvenile stage increased from 22.0 kg (average shell length 24.5 mm) to 75.5 kg (average shell length 42.5 mm) after 180 days. Feed conversion ratios increased slightly from 13.7 for the first 90 days to 16.3 thereafter. The shell growth rate of juvenile abalone between 24.5 mm and 34.8 mm was 3.4 mm month⁻¹, while for juveniles between 34.8 mm and 42.5 mm it was 2.6 mm month⁻¹. The total weight of abalone in the recirculating aquaculture system for the on-growing stage increased from 100.0 kg (average shell length 44.0 mm) to 433.3 kg (average shell length 72.7 mm) after 570 days. The feed conversion ratios for the first 173 days, the next 320 days, and the last 570 days were 19.6, 22.1, and 24.8, respectively. The growth rate of the average shell length during the on-growing period was 1.5 mm month⁻¹. Total ammonia nitrogen (TAN) concentrations were stabilized below 0.12 mg l⁻¹ in the juvenile recirculating system and 0.14 mg l⁻¹ in the on-growing recirculating system after conditioning of the biofilters.     

生物—电氧化法去除海水养殖循环水污染物

为提高海水养殖循环水处理效率,降低处理成本,本研究采用曝气生物滤器与电化学阳极氧化组合工艺,考察了不同阳极电势、进水氨氮和亚硝酸盐浓度下系统对氨氮及亚硝酸盐等污染物的去除效果,研究了微生物与工作电极之间的相互作用,并分析了电化学反应能耗。在水力停留时间为45 min、1.4 V阳极电压、进水氨氮和亚硝酸盐浓度分别为4.5和1.3 mg/L条件下,生物—电氧化法对氨氮去除率达88.8%,高出对照组7.6%,出水氨氮和亚硝酸盐浓度分别为0.5和0.9 mg/L,COD去除率为88.2%,高出对照组19.4%,平均能耗0.040 kWh/m~3,电极表面微生物生长对阳极电氧化过程有促进作用,微生物功能预测显示实验组硝化功能占比为0.03%,对照组为0.07%。研究表明,生物—电氧化法对海水养殖循环水的污染物有良好的去除效果,具有一定的发展应用潜力。     

不同氨氮和溶解氧条件下循环海水养殖系统生物滤池对氨氮、化学耗氧量及颗粒悬浮物的处理效果

为了建立优化的循环海水养殖系统,采用水质国标检测方法分析了珊瑚石生物滤池在不同氨氮和溶解氧(DO)负荷实验条件下对养殖废水中氨氮、化学耗氧量(COD)及颗粒悬浮物(SS)的处理效果。结果显示,进水氨氮浓度对出水氨氮(正相关)、COD(正相关)均有极显著的影响(P0.01),对SS处理效果影响不显著。当进水氨氮浓度为0.45~0.65 mg/L时,滤池对水体处理效果最优(氨氮平均清除率为82.1%±3.3%;COD平均清除率为7.1%±1.5%;SS平均清除率为5.8%±1.6%)。DO浓度对水体氨氮(负相关)和COD(负相关)处理效果的影响显著(P0.05),对SS处理效果影响不显著。DO浓度为5.0~7.0 mg/L时,水体处理效果最优(氨氮平均清除率为78.7%±3.5%;COD平均清除率为23.0%±5.3%;SS平均清除率为7.1%±2.0%)。因此,本实验环境下的循环海水养殖系统珊瑚石生物滤池在氨氮浓度为0.45~0.65 mg/L,DO浓度为5.0~7.0 mg/L时,对水体中的氨氮、COD、SS的综合处理效果最优。     

Mathematical Model of Channel Catfish Foodfish Production in Multiple-Batch

Market size requirements for catfish change periodically, and catfish farmers must adjust quickly. Data from catfish pond studies at the University of Arkansas at Pine Bluff (UAPB) were used to develop mathematical models of catfish foodfish in multiple-batch culture across a variety of management alternatives. Two different functional forms (Cobb–Douglas and a modified translog) were each developed into average and stochastic frontier models. Inefficiency terms were found to be non-significant in the frontier models, thus making the average and frontier models equivalent. In the average regression models, the modified translog form demonstrated superior statistical values as compared to the Cobb–Douglas form, but the latter resulted in lower prediction error and was validated with the Chow test when used to predict observations from commercial catfish farms. This approach appears to have merit from the perspective of its statistical properties, and represents a step towards development of a model that could be used for farm management purposes.     

基于径向基函数(RBF)神经网络的红鳍东方鲀体质量预测

基于表型性状预测红鳍东方鲀体重时,由于不同表型性状间的自相关、部分性状和体重之间的非线性关系以及线性回归方法自变量间的共线性,结果导致预测误差过大。为了解决这一问题,根据人工神经网络(Artificial Neural Networks, ANN)建模原理,采用径向基函数(Radial Basis Function, RBF) 模型,利用72个红鳍东方鲀样品的生长数据, 通过最近邻聚类算法, 构建了基于RBF神经网络的红鳍东方鲀体重预测模型,并采用线性回归检验法对所构建模型的可信度进行检验。结果发现,基于RBF神经网络预测模型的拟合优度为0.992,接近于1,而线性回归模型的拟合度为0.949。这表明: 基于RBF神经网络方法构建的预测模型消除了线性分析中自变量的共线性问题, 比线性回归预测模型的拟合度提高4.53%,预测精度高于线性回归方法。基于RBF神经网络体重预测模型的构建,为利用表型性状精确评估红鳍东方鲀的体重提供了一种新的方法。     

神经网络在养殖水质精准预测方面的研究进展

目前神经网络研究文献成果较多，虽然在水质精准预测方面起到了一定的参考，但由于文献缺少科学分类，使用率不高，导致学者难以找到研究切入点。针对这一问题，本文将神经网络方法在养殖区水质精准预测方面的文献按照海水和淡水两大领域进行分类，主要对每个领域所应用的预测模型从正反馈架构、循环架构和混合架构三个方向对海水时空序列文献进行分类研究和综述，发现混合架构模型的预测性能优于正反馈模型和循环架构模型，有利于提升不同深度水质预测模型的精度。另外，本文对基于神经网络方法的三维水质预测模型进行了初步探讨，发现学者的研究成果更多地集中在水表层和水中层的不同位置水质参数的变化方面，而神经网络方法对水表层水质预测精度比水中层和水深层水质预测精度高。