注記 |
A new method is presented for predicting the abundance of the rice stem-borer, Chilo suppressalis Walker, with objective of opening the way to solving the problem concerned. Some thirty years have elapsed since the predicting work of this pest borer of the rice plant was launched by the Japanese government in 1941, but much yet remains to be solved in the strict sense of the technique. A method newly developed consists of the following 5 steps in essence: (1) On the basis of a consistent relation between the percentage infested rice hills P and the mean number of infested stems per hill pt, i.e., P=1-e^-m, m=aμ^b, which was originally established for the second generation of the borer by Kono and Sugino ('58), a numerical table of μ available for the first generation of the borer was constructed anew in correspondence to the percentage infested rice hills P (P=0.01~0.99). (2) According to the current empirical limits for allowable economic degrees of infestation by this borer, that is, for-whether "needing control" or "not needing control", or eventually for the degrees of loss of yield, the possible economic degrees o£ infestation were classified into three: low(P being equal to, or lower than 5%, the upper limit for "not needing control"), moderate (P lying between the limits from 10% to 20%), and high (P being equal to, or higher) than 40%). Corresponding to each case, a set of an "acceptance" and a "rejection" equation of the sequential sampling test can be set up, respectively. (3) The total number of healthy larvae will be given by the mean number,(n-d)/t, of healthy larvae survived per infested stem, where n is the number of living larvae found in the total number (t) of infested stems sampled, d being the number of dead individuals during indoor storage of the collection. As regards the ability of dispersion of this stem-borer, the fact is known that one healthy individual larva attacks at least one stem until the middle and/or the last stage of each generation. Consequently, the average number, N, of healthy larvae survived per hill at these stages of each generation can be given by the product of (n-d)/t and μ, the mean number of infested stems per hill, which can be found in the numerical table Table II according to the value of P obtained from the sequential sampling test carried out at these stages. From these relations, it follows that the limits, N_i, for the values of N which correspond to three classes of economic degrees of infestation can be estimated. Then, the values of N_i will indicate the number of larvae per hill, representing the number of living larvae which are capable for the source of occurrence in the next generation. (4) Let N_xi be the limits for the values of N_i in the case where the mortality due to various environmental resistances including biological factors is subtracted. Then, the limits for the values of N_xi will be given by (figure) in correspondence to each economic degree of infestation by this stem-borer, where e is the average number of eggs to be deposited, m:f being the sex ratio of male to female, and w_k the k-th percentage compenent of environmental resistant factors concerned. These limits for the values of Nxi exactly give the limits for the values of μ_01, the mean numbers of infested stems per hill to be observed, and naturally from, the relation between P and μ, correspond to the values of P in the numerical table Table II. (5) Thus, let Pod be the percentage infested hills to be predicted towards the end of the next generation, then, in accordance with the economic degrees of infestation, the limits for the values of Pof can be given by the values of P in the numerical table Table II. Consequently, according as the values of P is smaller than 5%, or lie between 10% and 20%, and or larger than 40%, the limits for the values of P_0j also can be determined quite similarly. In like manner of inference, the probable percentage infested hills towards the end of any generation can be predicted by the decision, irrespective of the difference in generation, only if the numerical table (Table II) concerning P and μ is employed correctly, that is, against the prediction in the first generation, the table for the first generation use (Column Ist in Table II), and against the prediction in the second generation, the table for the second generation use (Column IInd in Table II) is employed, respectively. The advantage of this new method lies in the following three points: (1) A sequential sampling test which is effective in saving time, labour, and expenses was first introduced into the prediction of the abundance of the rice stem-borer and it is recommended that field surveys should be carried out of the percentage infested rice hills towards the end of a specified generation of the borer. The sequential sampling test here proposed is carried out on the three levels of infestation adopted empirically from the rates of loss of yield with the percentage infested hills as the criterion for decision. (2) On the basis of a consistent relation empirically verified between the percentage infested hills P and the mean number of the infested stems per hill μ, a numerical table of μ corresponding to P(=0.01~0.99) (Column Ist in Table II) was constructed anew for the first generation of the borer. Through this construction, the difference has been clarified between the case with the first generation and that with the second generation (Column IInd in Table II by Kono and Sugino, '58), and the quantitative relation between P and μ has been generalized still further, as a result of which the availability of the Table II has been enhanced greatly. In drawing final inductive inference of possible damages by this borer, special emphasis was placed on the effects of biological factors among various environmental resistances during the period between consecutive generations of the borer, such as the hibernating stage, the post-diapause stages, egg and larval stages arising from the first brood adults, and the adults themselves. (3) In addition to this, the way has been opened to the feasibility of elavating the current biological forecasting work to the economic prediction which is believed to be the ideal state of forecasting system concerned.
|
|