Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zmmu.msu.ru/files/images/spec/Russ%20Ent%20J/ent12_4%20441_445%20Morozov.pdf Дата изменения: Sun Jun 9 11:17:00 2013 Дата индексирования: Fri Feb 28 01:40:13 2014 Кодировка: Windows-1251 Поисковые слова: messenger

Russian Entomol. J. 12(4): 441445

ї RUSSIAN ENTOMOLOGICAL JOURNAL, 2003

Introducing entomophagous insects to control pests: prediction of target species density Интродукция насекомых-энтомофагов против вредителей: прогнозирование плотности популяции вида-мишени в случае интродукции насекомых Aleksey S. Morozov1, Svetlana V. Rytova1, Lynne C. Thompson А.С. Морозов1, С.В. Рытова1, Л.Ч. Томпсон2
1 1 2

2

The Moscow State Forestry University, 1 First Institutskaya Str. 1, Mytishy-5, Moscow region 14005 Russia. Московский Государственный Университет Леса, 1-я Институтская ул., 1, Мытищи-5, Московская область 141005 Россия. School of Forest Resources, P.O. Box 3468 - UAM, University of Arkansas, Moticello AR 71656-3468 USA.

KEY-WORDS: biological control, prediction, success. KЛЮЧЕВЫЕ СЛОВА: биологический контроль, прогноз, успех. ABSTRACT. It is second attempt prediction results of classical biological control. We retrospective assesed the perspective of the introduction of two species of parasites (Blepharipoda scutellata R.-D. and Parasetigena silvestris R.-D.) against Lymantria dispar L. The difference between predicted level of steady-state density of the host population and reality due to big errors estimated parameters (both for the parasites and the host). If for best investigated case the prediction is very approximate any assertion about absent perspective introduction parasites predators and competitor is false. РЕЗЮМЕ. Представлена попытка прогнозирования результатов классического биологического контроля. Учитывая опыт прошлого, мы оценивали перспективу интродукции двух видов тахин (Blepharipoda scutellataR.-D. иParasetigena silvestris R.-D.) против Lymantria dispar L. Разница между прогнозируемым результатом уровня стабилизации (особая точка) популяции хозяина и действительностью, обусловлена большими погрешностями оцениваемых параметров (как для популяции паразита, так и для популяции хозяина). Несмотря на то, что в данном случае имеется относительно богатая информация, результат все-таки получается очень приблизительным. Таким образом, всегда любое утверждение относительно отсутствия перспектив интродукции энтомофагов является ненадежным. moth [Howard and Fiske, 1911; Mc Gugan and Coppel, 1962]. Later on this methodology was used (with no 100 % success) in controlling numerous agricultural and forest pests. Success has been reached in 16% and partially in 58% of the cases [Hall et al., 1980]. It should be mentioned that the methodology came to practice prior to be properly justified, since the theory had not been adequately developed. Perhaps, only a part of it considered the issues climatic suitability [Coppel and Mertins,1977; Izhevsky,1990]was advanced. As for the biological basis of selecting parasitoid species for introduction, it has always been biased. In special literature one can find a list of requirments that introduced species should meet [Coppel, Mertins, 1977; Hall et al., 1980; Huffaker, Messenger, 1976; Izhevsky, 1990]. However, in practice low data quality or lack of information on some particular mechanisms of the introduction process did not slow down introductions. This approach was promoted, for instance, by Huffaker et al. [1971]. They wrote, in particular, that there is no need to study parasites and predators before introduction, since it is almost impossible to predict the effect of introduction. They proposed to introduce many parasitoid species so as to find out the best one. Probably, such an approach (supported to a certain extent by the majority of specialists) along with the improperly planned and made experiments as well as lack of detailed observations on parasites and predators, provided relatively low success of the introduction projects. There is huge number of population dynamics models for insects (literarute is not cited). However, any complete analysis of the introduction through even some promising models (see reviews Sharov, 1986; May, 1976) has not yet been accomplished. No evaluation of the behavior of the models under conditions similar to the natural ones was done [Varley, 1974; Hassell, Varley, 1973]. We think that the main reason of this abnor-

Introduction
Since the first half of the 20-th century introductions of predator and parasitoid insects aimed at pest control were widely applied. The primery impulse was made by large-scale introduction projects carried out in the USA and Canada mainly to control gypsy moth and brown-tail
Printed in 2004.


442

A.S. Morozov, S.V. Rytova, L.C. Thompson
4. Parasitoids emerge from all the hosts attacked; mortality of parasitoids at the span larvae in host imago is constant. As a basis of the model we use logistic equation discribing host-parasite interactions by Nicholson and Bailey [1935] and Hassell and Varley [1969]. The model parameters were estimated using SYSTAT (version 4.1). The following equation shows how host population density Nn+1 at the next generation depends on its density Nn at the current one: Nn+1=Nn*R*(1 Nn /K) *EXP(Qbl*Pbl,n1m1)* (1), EXP(Qpp*Ppp,n1m2) where R fundamental net reproductive rate the host population, K upper limit host population density, Pbl,n density of searching imago of B.scutellata of the current generation (number per 100 m2), Ppp,n density of searching imago of P. silvestris of the current generation (number per 100 m2), Qbl quest constant for B. scutellata, m1 mutual interference constant for B. scutellata, Qpp quest constant for P. silvestris, m2 mutual interference constant for P. silvestris. It is also assumed that P1 = N0*EGG*0.5*WBL*S1 (2), P2 = N0*EGG*0.5*WPP*S2 (3), where N0*EGG*0.5 is number of half-grown gypsy moth caterpillars accessible for the parasites in previous generation; 0.5 is survival of hosts at the stages egg-larvae of the 2nd instar; Wbl, Wpp proportion of hosts infected by B. scutellata and P. silvestris in previous generation, respectively; P1 and P2 post hunting density of populations of B. scutellata and P. silvestris, respectively. Lack of accurate data did not allow us to use variable EGG (mean number of eggs in the eggmasses in spring of the previous year). It was replaced by constant, namely by 350 (mean number of eggs in eggmasses). The constants S1 and S2 were calculated as follows: S1=0.45*0.5; S2=0.4*0.5, where 0.45 is survival of B. scutellata at the period larvae in host imago (the estimates given by Sisojevic [1975] and Zerova et al. [1989], were averaged); 0.4 is survival of P. silvestris at the period larvae in host imago (the estimates given by Sisojevic [1975] and Zerova et al. [1989], were averaged); 0.5 is proportion of females in parasitoid populations (actual data is not available). Dynamics of parasite populations was modeled as follows: PblN+1=Nn*S1*(1EXP(Qbl*Pbl1m1))*EXP(Qpp*Ppp1m2) (4) PppN+1=Nn*S2*(1EXP(Qbl*Ppp1m2))*EXP(Qbl*Ppp1m1) (5), where PblN+1, PppN+1 density of searching parasitoides of B. scutellata and P. silvestris of the next generaion, respectively. According to Hassell and Varley [1969]:

mal situation is lack of publications on detailed biological observations (presented in numerical form) as well as low quality of some biological research. Computerised expert systems are powerful tool for the forecast of results of classical biological control. Expert system is simply an instrument. It is a good tool if it is comfortable to use and using it gives good results. Such systems are useful for the education and the practical biological control. Prognosis is based on the retrospection analysis experience classical biological control and the modern theory of dynamics of population density. But mathematical approach is typically not use in practical biological control programs. Perhaps because of many entomologists are lacking in higher level mathematical knowledge. However, all the above does not mean that mathematical approach to the problem is impossible. Careful and critical review of published data as well as additional detailed observations on the majority of well-known pests and parasitoids can help provide valuable predictions using models.

Methods
Just to illuminate an approach to prediction of the results of introduction we chose gypsy moth (Lymantria dispar L.) one of the well-studied pests. Since detailed information on parasites and predators dynamics is needed to develop a model, we chose only two species parasitoids on which data were published in periodicals: 1) Blepharipoda scutellata R.-D. (=Sturmia scutellata R.-D., =Blepharipa scutellata R.-D., =B. pratensis Meig., =Blepharipoda pratensis Meig., =Herigia pratensisMeig.) (Diptera: Tachinidae); 2)Parasetigena sylvestris R.-D. (=Phorocera silvestris R.-D., =Ph. agilis R.-D., =Ph. segregata Rond.) (Diptera: Tachinidae). Both species are solitary internal parasites of half-grown and larger gypsy moth caterpillars. Gypsy moth are the most preferable host for them [Zerova et al., 1989]. Both species have been introduced to the USA and well established there [Clausen, 1956]. Three data sets (7, 11 and 14 year time series) on the host and both parasitoids referred to the native area were extracted from publications [Znamenski, Lyamtcev, 1990; Panina, 1985; Lyamtcev, 1985; Sisojevic, 1975; Yafaeva, 1963]. These data were used as a basis for modeling. The authors mentioned above have used different units for the estimates of the population density: number per 100 cluster of leafs [Znamenski, Lyamtcev, 1990; Panina, 1985; Lyamtcev, 1985], number per tree [Lyamtcev, 1985; Yafaeva, 1963], number per 100 m2 [Sisojevic, 1975]. All data were converted into number of individuals per 100 m2. The conversion was made using the methods proposed by Lyamtcev [1985] and regional tables of forest growth [Growth of the main ..., 1967]. Thus, we have only three rather short time series. Therefore, we need to simplify model structure as much as possible. The following assumption were used while constructing the model: 1. Populations of host and parasites are local (neither emigration nor immigration are supposed). 2. Both parasitoids act independently (i. e., probability of not getting attacked by both parasites is product of respective figures referred calculated for an each parasite separately). 3. All hosts are accessible for both parasitoids (no shelters exist).

Example of mathematical analysis of the introduction

QP1m =Ln

where Nn host population density at n-th generation, Nn W*Nn density of hosts remained uninfected, W proportion of hosts infected by both parasites. The estimation yielded the following values: B. scutellata: Qbl=0.213+0.022, m1=0.895+0.016, P. silvestris: Qpp=0.048+0.032, m2=0.596+0.116. Net fundamental reproductive rate (R) = 1.28 For the USA the data on the US populations was extracted from the article of Campbell [1967]. The role of the introduced parasites in 1910-1921 was considered as insignificant. The analysis of population dynamics of gypsy moth before introduction of parasites and predators was made using (7), logistic model: Nn+1=Nn*R*(1Nn /K) where R= 1.580+0.408 and K= 347.704+66.

Nn Nn - W * N

n

(6),


Introducing entomophagous insects to control pests: prediction of target species density

443

Results
Prediction of the results of introduction was made on the basis of model (1). We assumed that the searching ability of parasites (expressed by Q and m) did not change during the introduction. For mean case, forecasted mortality causedB. scutellata and P. silvestris was 33% and 5%, respectively. Ferguson et al. [1994] reported mortality 32%, and 26%, respectively. Mortality due this species is 24% and 27%, respectivily in homeland. The model runs simulating introduction showed sustainable increase of parasitoid population densities after temporary gradual decrease. Usually they also showed that host population density declined while reaching stability (N*). Temporary increase of host population density may occur at the beginning of the process, especially if the magnitude of R is comparatively high. Oscillations of parasitoid population density occur only in some cases, for instance if m>1. The results of the simulation are presented at Table 1 and Fig.1. While running the model, the following ranges of parameters were investigated. As lower limits of Q, m (Fig.1) the minimum possible values (mean value minus error) were used; the only exception was made for Qpp which lower limit was 0.001. The upper limits of the ranges were maximum possible values (mean value plus error). The ranges were devided into 10 (for Q and Q, m) and into 20 (for m and R). The runs were terminated when the difference between the current and subsequent values of the host density had become less than 0.0001. It usually took less than 50 iterations. The difference between levels of steady-state density of the host population can be of four orders due to big errors in the estimated parameters (both for the parasites and the host).

! & " $ & "

N
! & " $ & "

Q

m

Fig.1. Probably steady-state density (N*) of gypsy moth population past introduction parasites Рис. 1. Возможная точка динамического равновесия (устойчивая особая точка) популяции непарного шелкопряда после интродукции паразитоида.

Discussion
We know one attemt of prediction result of introduction parasites [Varley, Gradvell, 1968]. The result of introduction is remarkable similar to the prediction for biological control in Canada made by Varley and Gradvell in 1968.

Flareups of winter moth in Nova Scotia and British Columbia have occured at about 9- to 10-year intervals (although not neccessarily on oaks, and not at the high levels predicted [Roland, Embree, 1995]. The steady-state density of gypsy moth population (N*) is important characteristic of gypsy moth population dynamics. Most likely it is close to the mean population density. Analysis of the actual time-series showed that mean value of gypsy moth population density in the USA before application of classical biological control was about 58.0 eggmasses per 100 m2. It dropped down to approximately 3.9 eggmasses per 100 m2 [Cambell, 1967] upon completing the first programmes on introduction. If we also take into account

Table 1. Predicted steady-state level of density of gypsy moth and its parasites population. Таблица 1. Предсказанная точка динамического равновесия для популяции непарного шелкопряда и его паразитоидов

C ompanent of system L. dispar* Bl. scutelata** P. silvestris**

value of parametrs
mean Q1, Q2, m1, m2, R Q 1+s, Q 2+s, m1+s, m2+s, R Q 1s, Q 2s, m1s, m2s, R Q 1, Q2, m1, m2, R+s

12.4 36.7 12.9

10.5 19.9 7.5

17.7 123.2 22.7

50.3 123.0 87.9

* Mean number of eggmasses per 100 m2, ** Mean number of pupariaper 100 m2.


444

A.S. Morozov, S.V. Rytova, L.C. Thompson
200 pp. [in Russian] Hall R.W., Ehler L.E., Bisabri-Ershadi B. 1980. Rate of success in classical biological control of arthropods // Bull. Entomol. Soc. Amer. Vol.26. P.111114. Hassell M.P., Varley G.C. 1969. New inductive population model for insect parasites and its bearing on boilogical control // Nature. Vol.223. P.11331137. Hassell M.P., Varley G.C. 1973. Insect population ecology. An analytical approach. Oxford. 320 pp. Howard L.O., Fiske W.F. 1911. The importation into the United States of parasites of the gypsy moth and brown-tail moth // US. Bur. Ent. Bull.91. 344 pp. Huffaker C.B., Messenger P. S., De Bach P. 1971. The natural enemy component in natural control and the theory of biological control // C.B. Huffaker (ed.): J. Biol. Control. New York: Plenum press. P.1667. Huffaker C.B., Messenger P.S. 1976. Theory and practice of biological control. New York. 788 pp. Izhevsky S.S. 1990. [Introduction and application of entomophages. Moscow: Agropromizdat]. 222 pp. [in Russian] Lyamtcev N.I. 1985. [A study of the ecological and population parameters of prediction of numerical change in the gypsy moth population in the oak stands of wooded steppe]. Moscow: Doctor Thesis. 212 pp. [in Russian] May R.M. 1976. Models for two interacting populations // J. Theoret. Ecology. Sinauer associates INC. Massachusetts: Sundrland. P.78104. McGugan B.M., Coppel H.C. 1962. A review of the biological control attempts against insects and weeds in Canada // J. Biological Control Forest insect. Commonwealth Inst. Biol. Cont. Tech. Comm. No.2. P.35215. Korzuckhin M.D., Semevski F.N. 1993. [Synecology of Forest]. St.Petersburg: Gidrometeoizdat. 187 pp. [in Russian] Montgomery M.E. 1990. Role of site and insect variables in forecasting defoliation by the gypsy moth // Watt A.D., Leather S.R., Hunter M.D., Kidd N.A.C. Population dynamics of forest insects. P.7384. Nicholson A.J., Bailey V.A. 1935. The balance of animal populations // Proc. Zool. Soc. London.Vol.3. P.551598. Panina N.B. 1985. [Distribution and population dynamics of gypsy moth entomophages in the oak stands of the south-east of the European part of the USSR]. Moscow: Doctor Thesis. 20 pp. [in Russian] Price P.W. 1990. Evaluating the role of natural enemies in latent and eruptive species: new approaches in life table construction // Watt A.D., Leather S.R., Hunter M.D., Kidd N.A.C. Population dynamics of forest insects. P.221232. Roland J., Embree D. G. 1995. Biological control of the winter moth // Annu. Rev. Entomol. Vol.40. P.475492. Sharov A.A. 1986. The modelling of insect populations dynamics. Itogi naunmi i tekhniki [Advances in science and technology]. Entomologiya. Moscow: VINITI Publ. Vol.6. P.3115 [in Russian]. Sisojevic P. 1975. Dinamika populacije Tachina gubara u toku gubareve gradacije // Zastita bilja. Vol.26. No.132. P.97170. Varley G.C. 1974. Population dynamics and pest control // D.J. Price, M.E. Solomon (eds): Biology in pest and desease control. P.1527. Varley G.C. Gradwell G.R. 1968. Population models for the winter moth // T.R.E. Southwood (ed.): Insect Abundance. Oxford: Blackwell Sci. P.132142. Yafaeva Z.Sh. 1963. [Gypsy moth in Bashkiriya and the role of natural enemies in its limitation]. Moscow: Doctor Thesis. 210 pp. [in Russian] Zerova M.D., Kotenko A.G., Seregina L.Ya., Tolkanitc V.I. 1989. [Entomophages of Tortrix viridana L. and gypsy moth in the south-west of the European part of the USSR]. Kiev: Naukova dumka. 200 pp. [in Russian] Znamenski V.S., Lyamtcev N.I. 1990. [The peculiarities of the population dynamics of the gypsy moth in the multiple gradations of phytophagous insects] // Forest pest and desease management. Moscow: VNIILM. P.1121 [in Russian].

data by Bess [1961] the value droppes down more, to about 1.6 (log10-scale)1. It is lower than threshold density that is recommended to switch on chemical control, 6 eggmasses per 100 m2 , [Montgomery, 1990]. However, the problem of gypsy moth still remains partially due to inefficiency of the introduced parasites and predators [Price, 1990]. As it follows from our model analysis, introduction of two species of parasites (Blepharipoda scutellata and Parasetigena silvestris) may lead to stabilization of host population density. Its level (Fig. 1) is mainly higher than the threshold density recommended for switching on the control with chemicals. The predicted steadystate level of host population density is also much higher than the actual mean values. Partially it is caused by artificially low amount of parasitoid specied (2) used in the model. Actually, there are about 10 established parasitoids and predators [Clausen, 1956] affecting gypsy moth. We also have not paid any attention to the mammalian predators and birds. We have not made any comparison of the data on the role of the above two parasitoids after introduction with the predicted one. The reason is that we were not able to find parallel long time series for the USA, characterizing Blepharipoda, Parasetigena and gypsy moth. There is no doubt that this data was collected many times, because many authors have published the analysis of these data [Burgess, Crossman, 1929; Bess, 1961; Campbell, 1967]. The applicability of the methodology described above will become finally clear by accomplishing the similar analysis with respect to other species. However, we think that is could be already accepted for preliminary evaluation of the results of introduction. Of course, the accurate data on host and parasite population dynamics in the native area are needed to succeed.

References
Bess H.A. 1961. Population ecology of the gypsy moth Porthetria dispar L.(Lepidoptera: Lymantridae) // Conn. Agric. Expt. Sta. Bull.646. 43 pp. Burgess A.F., Crossman S.S. 1929. Imported insect enemies of the gypsy moth and brown-tail moth // USDA Tech. Bull.86. 148 pp. Campbell R.W. 1967. The analysis of numerical change in gypsy moth populations // Soc. Amer. Forest. Forest Sci. Monogr. Vol.15. 33 pp. Clausen C.P. 1956. Biological control of insect pests in the continental United States // USDA Tech. Bull.1139. 151 pp. Coppel H.C., Mertins J.W. 1977. Biological insect pest suppression, Berlin: Springer-Verlag. 314 pp. Ferguson C.S., Elkinton J.S., Gould J.R., Wallner W.E. 1994. Population regulation of gypsy moth (Lepidoptera: Lymantriidae) by Parasitoids: Does spatial density dependance lead to temporal density dependance? // J. Environ. Entomol. Vol.23. P.11551164. Growth of the main species of trees of the USSR. 1967. Moscow.
1 An adequacy of the estimation of mean population density of insect is needed in additional research, which was not the task of this work.


Introducing entomophagous insects to control pests: prediction of target species density

445

Appendix 1. Population density dynamics. Приложение 1. Динамика плотности популяции.

N1 15.84 5.74 4.36 0.96 0.10 0.15 0.23 0.21 0.53 3.29 6.96 7.91 40.00 33.50 16.00 1.00 1.50 2.50 10.50 12.00 13.00 5.30 341.90 248.10 2.90 0.06

N2 5.74 4.36 0.96 0.10 0.15 0.23 0.21 0.53 3.29 6.96 7.91 3.21 33.50 16.00 1.00 1.50 2.50 10.50 12.00 13.00 22.50 341.90 248.10 2.90 0.06 0.07

N0 39.62 15.84 5.74 4.36 0.96 0.10 0.15 0.23 0.21 0.53 3.29 6.96 15.00 40.00 33.50 16.00 1.00 1.50 2.50 10.50 12.00 0.30 5.30 341.90 248.10 2.90

W bl 0.422 0.594 0.561 0.526 0.150 0.125 0.086 0.100 0.1* 0.12* 0.559 0.535 0.030 0.330 0.180 0.001 0.0003 0.0003 0.0001 0.050 0.080* 0.049 0.122 0.260 0.397 0.047

Wc c 0.5 0.233 0.377 0.416 0.399 0.520 0.311 0.276 0.043 0.2* 0.3* 0.463 0.001 0.630 0.160 0.020 0.001 0.001 0.220 0.260 0.290 0.009 0.013 0.017 0.386 0.040

Egg 280.3 172.0 267.5 286.4 230.1 370.0 395.3 366.3 427.2 390.3 398.4 362.6 300.0* 314.0 359.0 309.0 221.0 396.0 380.0 405.0 331.0 350.0* 350.0* 350.0* 350.0* 350.0*

Comments and referenc e

Zname nski a nd Lyamtc ev, 1990 Lyamtc ev, 1985 Panina, 1985

Yafa eva, 1963

P. Sisojevic, 1975

* Estimated values; density is given in number of individuals per 100 m2; N1 density of eggmass in spring of Nn-th year, N2 density of egmass in spring of (Nn+1)-th year, N0 density of eggmass in spring of (Nn1)-th year, Wbl proportion of population parasitised by B. scutellata in (Nn1)-th year, Wcc proportion of population parasitised by P. silvestris in (Nn1)-th year, Egg number of eggs in eggmass in fall of (Nn2)-th year or in spring of (Nn1)-th year.