| Semanotus bifasciatus(Motschulsky) is one of important boring-insects,which was distributed in most provinces in China.It mainly damages such trees as Platycladus orientalis,Sabina chinensis and Chamaecyparis.At present,the forecast system of Semanotus bifasciatus is unable to cope with its disaster inadequately focuses on occurrence period.So we need to enhance the forecast system.This study used geo-statistics to measure the spatial pattern of the larvae,and explored the relationship between the intensity of photosynthesis and the damaged extent of Platycladus orientalis.Both are aimed at establishing a new forecasting method.We attempted to use artificial neural network technology to forecast the extent of the occurrence and improved attracting and killing technology to avoid the adult which in wood bait trapping for monitoring may disperse into live woods.The main results are as follows:1.We surveied the gummosis hole[represents Semanotus bifasciatus larva which harms in tree] on sunny slope and shady slope,and use geo-statistics methods to study the larva spatial distribution pattern.The results show:It is different in larvae density of Semanotus bifasciatus between sunny slope and shady slope.The distribution patten of larvae showed spatial aggregation in both slopes.The density of larvae in shady slope is lower than in sunny slope.The reason is that Semanotus bifasciatus usually oviposit on same tree and prefer to hazard in sunny slope.After the density of larva amount reaches to some extent,the larva shows a trend to spread around lack of food.2.It is difficult to observe the larvae after it bores into the tree.We studied the photosynthesis difference between the trees damaged at various extent and the healthy trees in order to find a new method for determining whether a tree is damaged or not in earlier phase.The results shot that:The levels of photosynthesis and water-stress are different between the damaged trees and the health trees. The process curve of photosynthesis in health trees was "two peaks" curve while the diurnal variation of the net photosynthetic rate and the transpiration rate was the "single peak" curve in damaged trees, which illustrate that the Photosynthesis of trees get weaken.The reason for "single peak" curve may be due to the lower enzyme activity related to photosynthesis in the damaged trees.Therefore,it is feasible to ascertain whether a tree is damaged or not by detecting changes on photosynthesis.The process curve of photosynthesis in Platycladus orientalis with 25%withered leaf and with 50%withered leaf also appears "two peaks" curve.But it is differert with health tree,because it is not caused by Semanotus bifasciatus.3.In order to establish the forecast model,the BP neural network was used.In the study,we made an investigation in the Ming tombs forest farm and processed the investigated data to establish the model.These data contained aspect,slope,height,diameter,and canopy density.By using the neural network toolbox in Matlab,but we didn't found the way to complete the learning process even if adjusted many times.We conclude that the learning process can not be completed probably because the valid data is too small.4.An indoor trial on the screening of insecticide was carried in the forest protection lab of Beijing Forestry University in order to improve attracting and killing technology with trap log.The wood with chlorpyrifos sprinkled on its surface had the best effect.Field trial was carried on Mushi Farm in Tengzhou city in Shandong province by using three insecticides:beta-cypermethrin,compound of avermectin and chlorpyrifos,profenofos.The results show that:Beta-cypermethrin diluted 1000 times and 2000 times has the best control effects.The insecticides would not reduce the attractive effect of trap log.When economic factors are taken into account,it is recommended that the trap log should be sprinkled by beta-cypermethrin diluted 2000 times and vertically placed.
|
PDF Full Text Request