CN102550454A - Method for predicting cryptocaryoniosis in Larimichthys crocea - Google Patents

Abstract

A method for predicting cryptocaryoniosis in Larimichthys crocea relates to Larimichthys crocea, and belongs to a method of accurately predicting a water environmental factor at a certain future point-in-time to achieve advanced prediction on cryptocaryoniosis in Larimichthys crocea. The method includes: plugging a predicted value of the obtained water environmental factor in a mathematical discriminatory model for cryptocaryoniosis based on the water environmental factor; calculating a predicted value of the cryptocaryoniosis to generate the advanced prediction on the cryptocaryoniosis in Larimichthys crocea; and obtaining advanced prediction on the cryptocaryoniosis in Larimichthys crocea. The method is simple and convenient and capable of reflecting anural variation, parameters of the water environmental factor can be obtained according to a series of fitting equations only by providing month, the parameters can be calculated with a disease discriminatory model to finally obtain the predicted value of the cryptocaryoniosis, and accordingly, the method is quite convenient in use. For the specific cryptocaryoniosis, the method is convenient for farmers to timely take measures to control the disease and prevent and alleviate disasters.

Description

A Prediction Method of Cryptocaryoniasis Stimulated by Large Yellow Croaker

technical field

The invention relates to a large yellow croaker, in particular to a method for predicting cryptocytoniasis stimulated by large yellow croakers.

Background technique

Large yellow croaker (Pseudosciaena crocea) is the fish with the largest output of a single species in marine cage culture in my country, with an annual output of about 70,000 tons. However, in recent years, stimulating the outbreak and spread of Cryptocaryon irritans disease has seriously threatened the sustainable development of large yellow croaker aquaculture industry. According to statistics, since 2005, large yellow croakers in the Sanduao breeding area of Ningde City, Fujian Province have caused direct economic losses of more than 300 million yuan each year due to secondary infections caused by parasites and bacteria. In 2008, the Ministry of Agriculture has listed Cryptocaryoniasis as a second-class animal disease in the new version of "List of Types I, II, and III Animal Diseases".

Once the cryptocystosis stimulated by large yellow croaker breaks out, the effect of using drugs to control it is very limited, especially when the stimulated Cryptocaryonia invades the body surface and gills to form trophozoites, it is even more difficult to effectively kill them with chemical drugs. In addition, serious problems such as water pollution and permanent residues caused by excessive use of chemical pesticides have brought new environmental problems, which ultimately affect the quality and safety of aquatic products and the sustainable and healthy development of marine aquaculture. In the long-term breeding practice, farmers have summed up some preventive and effective disease prevention measures, such as timely thinning the breeding density or moving them to deep water areas with better water flow in the early stage of disease outbreaks, and reducing the amount of fresh and miscellaneous baits Or add some immune enhancers, etc., and have received a better preventive effect. It can be seen that timely and active preventive measures before the onset are more important than remedial measures for drug prevention and treatment after the onset. Therefore, it is very necessary to establish an early prediction technology for the occurrence of Cryptocaryoniasis stimulated by large yellow croaker and to guide the early prevention and control of the disease.

my country already has a prediction method for large yellow croaker breeding diseases (all diseases) based on transparency, temperature and average wind speed (see Chinese patent: 200710068792.6). However, there are various types of large yellow croaker breeding diseases, and their epidemic characteristics are different. In particular, the effects of environmental factors involved in the concurrent occurrence of multiple diseases at the same time point are very complex, and the main effect factors affecting different diseases are not completely the same. According to specific water quality factors It is difficult to predict all diseases.

The occurrence of cryptocystosis stimulated by large yellow croaker is closely related to the complex aquaculture ecosystem in the main production areas of large yellow croaker in my country. The relationship with water environment factors was qualitatively described. On this basis, if a prediction technology based on key water environment factors can be established to predict and discriminate specific Cryptocaryonia stimuli, it will be of great significance for the prevention and mitigation of this malignant disease. However, there is no research report on the prediction technology of water environment factors related to cryptocystosis in my country, and there is no mature method for predicting cryptocystosis in large yellow croaker.

Contents of the invention

The object of the present invention is to provide a method for predicting cryptocytosis of large yellow croaker that can accurately predict water quality environmental factors at a certain time point in the future and realize the prediction of cryptocystosis of large yellow croaker in advance.

The present invention comprises the following steps:

1) Construction of water environment factor prediction equation

According to the historical data of water environmental factors in different months, a variety of time series models are used to analyze the changing laws and trends of water environmental factors, and the fitting curves of water temperature, dissolved oxygen and ammonia nitrogen values to months are respectively constructed to compare the fitting effects and select the best results from them. A good model establishes a fitting equation, and its regression equations are:

(1) Water temperature:

f(x)＝20.93-2.835*cos(0.5276*x)-8.043*sin(0.5276*x)

Where f(x) is the water temperature and x is the month;

(2) Dissolved oxygen:

f(x)＝7.083+-0.2234*cos(0.02547*x)+0.179*sin(0.02547*x)+0.9994*cos(2*0.02547*x)+0.9807*sin(2*0.02547*x)+0.6179* cos(3*0.02547*x)+0.1761*sin(3*0.02547*x)+0.2184*cos(4*0.02547*x)+0.002575*sin(4*0.02547*x)

Where f(x) is dissolved oxygen and x is month;

(3) Ammonia nitrogen:

y(t)=0.182y(t-1)+0.778y(t-2)+e(t)

Among them, y(t) is the actual value of the ammonia nitrogen value of the current period, y(t-1) is the ammonia nitrogen value of the previous period, y(t-n) is the ammonia nitrogen value of the previous n periods, and e(t) is the error value of the current period, Then the actual value of ammonia nitrogen in the current period is the predicted value of ammonia nitrogen in the current period plus the error value;

2) Prediction of water environment factors in advance

Calculate the predicted values of water temperature, dissolved oxygen and ammonia nitrogen at a specific time point in the future according to the established fitting equation;

3) Stimulate early identification of Cryptocaryon morbidity

Collect historical months, water temperature, dissolved oxygen and ammonia nitrogen 4 factors and the corresponding stimulus cryptocystosis severity data, use the random forest program package loaded in the R software environment to collect historical months, water temperature, dissolved oxygen and ammonia nitrogen 4 factors and the corresponding severity level data of Cryptocaryoniasis stimuli were analyzed, and a mathematical model was established to distinguish the disease level from the 4 factors of month, water temperature, dissolved oxygen and ammonia nitrogen. Predict the water temperature, dissolved oxygen and ammonia nitrogen values during the period, and then substitute the predicted values into the established disease discrimination model to predict the stimulation of cryptocystosis by large yellow croaker, and make corresponding interpretations to judge the possibility of stimulation of cryptocystosis According to the occurrence situation, appropriate prevention and control measures should be taken to reduce the loss due to the disease; the accuracy evaluation of the obtained predicted value can be compared with the actual occurrence of cryptocystosis stimulated by large yellow croaker.

The technical scheme of the present invention is to firstly substitute the predicted value of the obtained water environment factor into the mathematical discriminant model of stimulating Cryptocaryoniasis based on water environment factors, calculate the predicted value of the disease, and form the prediction value of large yellow croaker stimulating Cryptocaryoniasis. Advanced prediction technology, so as to realize the early prediction of the occurrence of cryptocystosis stimulated by large yellow croaker.

Compared with the prior art, the present invention has the following advantages:

1) Simple and convenient. The water quality prediction technology established by the invention based on the historical database reflects the annual change rule, and the parameters of the water environment factors can be obtained according to a series of fitting equations only by providing the month. Combined with the disease discrimination model for calculation, the predicted value of the disease is finally obtained, which is very convenient to use.

2) Prediction in advance. The present invention provides a technology for predicting cryptocystosis of large yellow croaker in advance for the specific cryptocystosis irritating for the first time, which is convenient for farmers to take early measures for disease prevention and control and disaster prevention and mitigation.

Description of drawings

Figure 1 is a trend fitting diagram of water temperature with respect to time. In Figure 1, the abscissa is time (month), and the ordinate is water temperature (°C); the mark "◆" is the actual value, and "-" is the forecast curve.

Figure 2 is a trend fitting graph of dissolved oxygen with respect to time. In Figure 2, the abscissa is time (month), and the ordinate is dissolved oxygen (mg/l); the mark "◆" is the actual value, and "-" is the forecast curve.

Figure 3 is the autocorrelation analysis results of the ammonia nitrogen value. In Figure 3, the abscissa is the lag order Lag, and the ordinate is the sample correlation Sample Autocorrelation.

Figure 4 is a trend fitting diagram of ammonia nitrogen with respect to time. In Figure 4, the abscissa is time (month), and the ordinate is NH4 ⁺ -N (mg/l); the mark "△" is the actual value, "+" is the predicted value, and "~" is the interpolation line.

Detailed ways

The following embodiments will further illustrate the present invention in conjunction with the accompanying drawings.

Example 1: Establishment of water environment factor prediction technology.

The present invention first collects the monthly monitoring data of water temperature, dissolved oxygen and ammonia nitrogen value of 5 different stations in the large yellow croaker breeding sea area of Sandu Bay, Ningde, Fujian (the water quality data comes from the East China Sea Sub-bureau Fujian Ocean Environment Monitoring Center Station of the State Oceanic Administration, and the disease data From the Department of Disease Prevention and Control of Fujian Marine Aquaculture Technology Extension General Station, data collection and analysis are in accordance with national standards), and a regression equation reflecting its changing trend and historical law was established. Among them, the months are recorded as 1-65 according to January 2005 to May 2010; the water temperature, dissolved oxygen, and ammonia nitrogen values at different stations are respectively averaged, and the trends and curve fitting effects of various methods are compared. Get the best fit equation.

1) Water temperature prediction technology

The water temperature changes with the seasons and has obvious periodicity. The present invention selects the Fourier function to fit the water temperature with the monthly variation trend, and the results are as follows:

f(x)＝20.93-2.835*cos(0.5276*x)-8.043*sin(0.5276*x)

Where f(x) is the water temperature and x is the month.

The R ² of the fitting curve is 0.9528, which shows that the fitting effect of the first-order Fourier function on the variation of water temperature with the month is good, and the fitting diagram is shown in Fig. 1 .

2) Prediction technology of dissolved oxygen

Dissolved oxygen is closely related to environmental factors such as the breeding mode of large yellow croaker, tidal rhythm, and annual changes of phytoplankton, and is an important factor that stimulates the occurrence of cryptocystosis. The present invention selects the Fourier function to fit the trend of dissolved oxygen with the month, and the results are as follows:

f(x)＝7.083+-0.2234*cos(0.02547*x)+0.179*sin(0.02547*x)+0.9994*cos(2*0.02547*x)+0.9807*sin(2*0.02547*x)+0.6179* cos(3*0.02547*x)+0.1761*sin(3*0.02547*x)+0.2184*cos(4*0.02547*x)+0.002575*sin(4*0.02547*x)

Where f(x) is dissolved oxygen and x is month.

The R ² of the fitting curve is 0.9369, which shows that the 4th-order Fourier function has a good effect on fitting the change of dissolved oxygen with the month. The fitting diagram is shown in Fig. 2 .

3) Prediction technology of ammonia nitrogen

Ammonia nitrogen is an important factor to measure the quality of water environment, and it is also one of the main factors that affect and stimulate the occurrence of Cryptocaryoniasis. The present invention selects an autoregressive model to simulate the changing trend of ammonia nitrogen. Through the results of ACF analysis (see Figure 3), a 2-order autoregressive model was selected, and the results are as follows:

y(t)=0.182y(t-1)+0.778y(t-2)+e(t)

Among them, y(t) is the actual value of the ammonia nitrogen value of the current period, y(t-1) is the ammonia nitrogen value of the previous period, y(t-n) is the ammonia nitrogen value of the previous n periods, and e(t) is the error value of the current period, Then the actual value of ammonia nitrogen in the current period is the predicted value of ammonia nitrogen in the current period plus the error value. Therefore, the predicted value of ammonia nitrogen in the current period y(t)=0.182y(t-1)+0.778y(t-2);

Loss function of this model＝0.000155492; FPE＝0.000180371

It shows that the autoregressive model has a good simulation effect, and the prediction simulation is shown in Figure 4.

Embodiment 2 Accuracy Analysis of Water Environment Factor Prediction Technology

According to the fitting equation established in Example 1, the water temperature, dissolved oxygen and ammonia nitrogen values in July of a certain year were predicted. Table 1 compares the predicted value with the actual value. It can be seen from Table 1 that the average relative error between the predicted value and the actual value of each factor is within 10%, indicating that the water quality prediction effect is good.

Table 1 Comparison of predicted and actual values of water quality factors in July of a certain year

factor actual value Predictive value Relative error(%) Water temperature (℃) 28.0 28.6 2.1 Dissolved oxygen (mg/l) 5.86 6.04 3.1 Ammonia nitrogen (mg/l) 0.030 0.034 9.7

Note: relative error = (predicted value - actual value) / actual value * 100%

Example 3 stimulates the advance prediction of Cryptocaryon morbidity

Collect the historical month, water temperature, dissolved oxygen and ammonia nitrogen 4 factors and the corresponding severity level data of cryptocystosis, use the random forest program package loaded in the R software environment to analyze the above data, and establish the month, water temperature, dissolved oxygen and A mathematical model for distinguishing disease grades with 4 factors of ammonia nitrogen.

The predicted value of water temperature in July of a certain year is 28.6°C, the predicted value of dissolved oxygen is 6.04 mg/l, and the predicted value of ammonia nitrogen is 0.034 mg/l. Substituting it into the established disease discrimination model can predict the stimulus in July of a certain year. The incidence of cryptocystosis.

After calculation, the degree of stimulating cryptocystosis in July of a certain year is level 2, that is to say, it is in a state of a small amount of disease, and the areas with high breeding density can take appropriate measures such as translocation and thinning to avoid the disease.

The prediction results of Cryptocaryonia stimuli are in line with the actual situation, which shows that the early warning technology has high accuracy.

Claims (1)

1. a method for predicting cryptocytoniasis stimulated by large yellow croaker, is characterized in that comprising the following steps: 1) Construction of water environment factor prediction equation According to the historical data of water environmental factors in different months, a variety of time series models are used to analyze the changing laws and trends of water environmental factors, and the fitting curves of water temperature, dissolved oxygen and ammonia nitrogen values to months are respectively constructed to compare the fitting effects and select the best results from them. A good model establishes a fitting equation, and its regression equations are: (1) Water temperature: f(x)＝20.93-2.835*cos(0.5276*x)-8.043*sin(0.5276*x) Where f(x) is the water temperature and x is the month; (2) Dissolved oxygen: f(x)＝7.083+-0.2234*cos(0.02547*x)+0.179*sin(0.02547*x)+0.9994*cos(2*0.02547*x)+0.9807*sin(2*0.02547*x)+0.6179* cos(3*0.02547*x)+0.1761*sin(3*0.02547*x)+0.2184*cos(4*0.02547*x)+0.002575*sin(4*0.02547*x) Where f(x) is dissolved oxygen and x is month; (3) Ammonia nitrogen: y(t)=0.182y(t-1)+0.778y(t-2)+e(t) Among them, y(t) is the actual value of the ammonia nitrogen value of the current period, y(t-1) is the ammonia nitrogen value of the previous period, y(t-n) is the ammonia nitrogen value of the previous n periods, and e(t) is the error value of the current period, Then the actual value of ammonia nitrogen in the current period is the predicted value of ammonia nitrogen in the current period plus the error value; 2) Prediction of water environment factors in advance Calculate the predicted values of water temperature, dissolved oxygen and ammonia nitrogen at a specific time point in the future according to the established fitting equation; 3) Stimulate early identification of Cryptocaryon morbidity Collect historical months, water temperature, dissolved oxygen and ammonia nitrogen 4 factors and the corresponding stimulus cryptocystosis severity data, use the random forest program package loaded in the R software environment to collect historical months, water temperature, dissolved oxygen and ammonia nitrogen 4 factors and the corresponding severity level data of Cryptocaryoniasis stimuli were analyzed, and a mathematical model was established to distinguish the disease level from the 4 factors of month, water temperature, dissolved oxygen and ammonia nitrogen. Predict the water temperature, dissolved oxygen and ammonia nitrogen values during the period, and then substitute the predicted values into the established disease discrimination model to predict the stimulation of cryptocystosis by large yellow croaker, and make corresponding interpretations to judge the possibility of stimulation of cryptocystosis According to the occurrence situation, appropriate prevention and control measures should be taken to reduce the loss due to the disease; the accuracy evaluation of the obtained predicted value was compared with the actual occurrence of cryptocystosis stimulated by large yellow croaker.

Images