Experimental & Applied carly, 12 (1991) 259-218 29 Eevier Scene Polishes BV Amsterdam Environmental factors affecting the life-tables of Tetranychus urticae (Acari: Tetranychidae). IIL. Host-plant nutrition 1B, Wermetinger, J.J, Oertli? and J. Baumgiriner +s Federal Isie of Fret, Sr dnd Landicape Research Deparment Landscape Bec: "Zurcher 11, 3905 Bos Switand "ode tte of Tchology ETH, Daren’ of Pat Sines, Crop Sie Research Son "Ethikon 4315 Lindau icon Sierland Federal tit of Teco ETH, Dparimen of Pat Stee, Phyomediine, Clauss 11,8092 rch Stead (Aceped 6 Mare 1991) \Wermeings 8, Oe 1J.and Bauder, 191 Eavrnmental factors testing te ieee i Ternchas eae (Act Tengehida) Hesse tion, Exp 9p ara 1 see Fis the rte fhe a wo dese sutton ees of ost plants on spider mits is bey seviwed Second experiments are Ssribed th sjcte micro ropagse spe 0 ‘Gur ent ies of ech macronutrient NP ad K Spier mies (Tevanetas wea Koch) fEoling eas of tee pat mre chee fore ceveoptetal ine, ag produton and Ionut Pan anaes rveed tthe concentationo, Pa Kcomesponded oe especie Treatments The coment pheno cmnouncs nthe lees ncaa wits Nand dete Io {Hen experimen spider prima! developer ate nd potion ate wae Yoh Bo {fey conte hea NO, xandity te poly coveted wie N ard caoohtate ‘Smet oe eves end eavely with he pheno sntent Longe) of he two spoed sie ‘rte nao signa ate by any taint The Kenprent ded ony inorder erin plant cotets a el ab spider ley om net ats, wee constrciea s stasalyaayzd byte Jackie ch nique, Toe hetable nays showed ral delet he nasa of ata eens (a) ‘tit Nand fein. Wiha espetments poled. scary crite ea ad partie. Ul othe content of pheno compounds inte ees, igen soaps tad ihe et stint Fnucnecon me population grow ina ange of LSS ea Nese bya ator of, the rum mutisions pe enero fo) by 11 and he doing te of he population vas song fold on seveely Nice leaves. INTRODUCTION ‘The growth of a spider-mite population is modified within a genetically defined spectrum by various environmental factors. Among the environmen- (0168-8162/91/803.50 © 1991 Elsevier Science Publishers LV, Allright reserved.260 a woratuncen erat tal influences are climate, food availability and quality, predation, and inter- and intraspecific competition. The influence of temperature and host-plant condition as affected by water stress Eave been discussed in previous papers of this three-part series (Wermelinger etal, 1990a,b). Under standard con- ditions, the developmental rate of the two-spotted spider mite, Tetranyehus urticae Koch, and its intrinsic rate of satura increase (rq) were found to be linearly related to temperature in the range between 15°C and 35°C. Gener- ation time gradually declined with increasing temperatures in Unis range ‘Temperature increase also turned out tobe the most important effect of host. plant water stress; the shortage of wacer supply caused leaf temperatures 0 rise, thus decreasing developmental time. The performance of a mite popt- lation under water stress was found to be the result of a balance between ac- celerated development due to higher leaf temperatures and reduced reproduc- tion due to inferior food quality The significance of food quality as a further component of environmental influences had been recognized early, and there exist a numberof reports on the influence of plant nutrition on spider-mite population development. In ‘most studies, mite densities are related to fertilizer treatments orto the m- tient content ofthe plant. By way of contrast we seek, in our work, a deeper understanding of this relationship by analyzing the effect of the nutritional status on the mite life-table parameters and their variances, rather than on population densities. This is made possible by recent advances in demo- sraphic theory and methodology (Krebs, 1989). In practice, the advent of ‘computers enabled the statistical analysis of life-tables Individual single life-history parameters (such as developmental time, fe- cundity, survival) of the spider mites as well as theit life-table parameters (such as net reproductive rat, intrinss rate of increase, generation time) are subjected to statistical analyses. The data are related to the physiological state of the plant expressed as contents of racroelements and other nutritionally important compounds. Hereafter, a ¢istinction will be made between two kinds of stress, namely deficiency and excess nutrient supply tothe host plant, LITERATURE REVIEW The literature on the nutritional effects of food plants on insects has been ‘suramarized in elaborate detail by Dale (1988). The specific role of nitrogen as discussed by Mattson (1980). Watson (1964) and Van de Vrie etal (1972) reviewed the effects of plant mineral nutrition as it concemed spider rites. Van der Geest (1985) summarized the studies that tied to determine ‘the nutritional requirements to rear spider mites on artificial diets. A number of further studies have since been made on how plant nutrition affects tetran- ychids, and the present synopsis focuses on papers published during the last ‘wo decades,NP. K nutrition ‘Nitrogen. Most attention has been given tothe effect of host-plant N on spider mites. Many studies investigated its influence on population densities: nitrate fertilization of apple trees led to higher population numbers of Panonyehus ulmi (Koch), depending on the date of application ( Van der Vrie and Delver, 1979). High contents of carbohydrates anc insoluble N compounds in the N- deficient treatment were correlated with higher multiplication rates of the same mite species (Fritzsche et al., 1980). On the other hand, carbohydrate contents of different plant lines had a negative effect on Oligon}chus pratensis (Banks), and the N concentrations in the leaves were not clearly related to ‘mite densities (Perring et a, 1983). Mellors and Propts (1983) found no significant effects of different N fertilization on T. urtieae in combination ‘with varying soil moisture. A larger population of Eotetranychus hicoriae ‘McGregor on N-ferilized pecan was attributed to enhanced reproduction and. not toa shorter life-cycle (Jackson and Hunter, 1983), The N content of leaves, of various cucurbit varieties positively influenced populations of T- neacale- donicus Andre (Sharma and Pande, 1986). Nitrogen fertilization of sorehum and corn only slightly affected 0. pratensis densities (Archer et al., 1988). (Other studies concerned the effects of N on life-history parameters: N fertil- ization was found to trigger higher levels of amino acids in some vegetables and ornamental plants, and some ofthe essential amino acids were correlated with elevated fecundity of 7. urticae (Tulisalo, 1971). Suski and Badowska (1975) reported that increasing N doses shortened developmental time and promoted survival of 7: urticae. The fecundity ofthis species was reported to be linearly correlated with N, sugar and water content of apple leaves (Wer- melinger et al., 1985). Fecundity and developmental time of T. pacficus (McGregor) were favorably influenced by fertilizer-induced increase of leat N (Wilson et al, 1988). Respiration rate and sex-ratio of T: urtcae were increased by elevated leaf N( Wermelinger, 1989; Wermelinger and Delucchi, 1990). The few studies on the N effect on soider-mite life-table parameters as overall indices of mite population performance unanimously found a stimu lating effect of N fertilization and leaf N concentration on the net reproduc- tive rate, Ry (Suski and Badowska, 1973) and on the intrinsic rate of in- crease, rq (Jesiotr et al., 1979) of T. urteae. Also, Yaninek et al. (1989) attributed enhanced r Values of Mononychellus anajoa (Bondar) on young ‘cassava leaves to higher leaf N concentration. In general, @ positive effect of \N fertilization and/or N content ofthe plant can be postulated for life-history parameters of spider mites. Phosphorus. The effects of P oa spider mites are less uniform, and fewer stud ies have been made concerning this element. Suski and Badowska (1975) {ound high P supply to enhance the survival of P. ulmi, Higher values of car-‘bohydrates and insoluble N compounds in apple plants, induced by a P-defi~ cient treatment, were positively correlated with higher population growth of T-urticae (Fritesche et a., 1980). TheP content of leaves of various cucurbit varieties, onthe other hand, could not be related to the population size of 7 neocaledonicus Andre (Sharma and Pande, 1986). Phosphorus fertilization ‘of sorghum and corn showed only a sight influence on O. pratensis (Archer etal, 1988). Potassium. Deficiency of K caused higher levels of amino acids in vegetables and ornamental plants, and an elevated fecundity of T- urtieae (Tulisalo, 1971). Similarly, more eiophyids (Erlophyes sheldoni Ewing) developed on plants in K-free nutrient solutions than in standard solutions (Sternlicht et al, 1975). In addition, Jesitr etal. (1979) reported that good K supply in- ‘duced higher immature mortality of the two-spotted spider mite. Corre- ‘ponding to these findings, the rate of population increase (7x) of T. urticae ‘vas found tobe negatively correlated with the K content of bean plants (Suski ‘and Badowska, 1975). This may be duc to higher values of carbohydrates and soluble N compounds at K-deficiency which result in higher multiplication rates (Fritzsche et al, 1980). On the other hand, the K content of leaves of various cucurbit varieties was reported to positively influence populations of T. neocaledonicus Andre (Sharma ané Pande, 1986) [Effects of ‘good’ and ‘poor leaves cn T: cinnabavinus (Boisduval), as de- scribed by Weensch and Young (1978, resulted from crowding and previous feeding rather than from plant nutrition. The sex-ratio of 7: urticae was higher ‘on leaves of poor’ than of good’ quality (Kondo and Takafuji, 1982), Other nutritional components ‘Only a small number of articles coacemed trace elements and secondary plant metabolites. Oil plant seedlings receiving litle boron experienced a more intense attack of T. plercei McGregor than did well-supplied seedlings (Ra- jaratnam and Law Ing Hock, 1975). Manganese application together with ammonium reduced 7” arabicus Attiah populations, but the same application ‘without ammonium gave contrary resilts (Abou-Awad, 1981). Primitive races of cotton that were almost immune to T: uricae had con- centrations of tannins (phenols) of up to 20% higher (Lane and Schuster, 1981 ). Oligonychus pratensis was found tobe insensitive to tannins (Perring etal, 1982), but the phenol content of peppermint leaves exerted a signifi- cant impact on 7: wrticae (Larson and Berry, 1984). On young leaves with high phenol levels, they produced the lowest number of eggs, the develop- ‘mental time ofthe immatures increased and their dispersal was enhanced. A similar linear relationship between phenol content and fecundity was re- ported by Wermelinger etal. (1985).Concluding remarks Discrepancies between the results of the reviewed studies are partly due to different ranges of fertilization, These are difficult to compare because of df- ferent methodologies (field, greenhouse, soil nutrient solutions) and plant species, Often, spider-mite development isrelated tothe variation of one nu- tient in the plant, assuming that the other components remain unchanged. However, herbivore performance is more likely the response to the overall physiological state of the plant, Very few studies related lifestable analyses of T-urticae to the N, P, K composition ofthe plant, and none related life-tables te secondary plant metabolites or made life-table statistics. The present work should help to fill these gaps in our knowledge. MATERIAL AND METHODS Experiments All plant material originated from one seed of the apple variety ‘Glocken- apfel’ which was shoot-propagated and rooted by means of tissue cultures (Wermelinger, 1985). The plantlets were transferred when 5.em high to low- concentration nutrient solutions under non-sterile conditions. At a height of 10.em they were transplanted to the nutrient solutions of the respective treat- ‘ments, composition of which was based on a Hoagland solution (Epstein, 1972), the standards (1 N, 1 P and 1 K) containing 210 ppm N, 31 ppm P ‘and 235 ppm K at pH 6.0, Based on the standards, concentrations of each of IN, P and K were varied by factors of 5, 1/5, and 1/25, thereby keeping the other two macronutrients and the micronutrients constant. Hence, the four treatments for N, for example, are referred to as 5N, 1 N, 0.2 N and 0.04 N, For each nutrient N, P and K, a separate experiment was carried out. Each ‘reatment consisted of eight plants in two containers holding 30 | of nutrient solution which was aerated and mixed by compressed air. The solutions were renewed monthly. At the end ofthe experinent the apple tres were about 1.5 mhigh ‘Six weeks after commencing the experiments, leaf disks (o 4 em) were punched out of young leaves, on which females ofthe two-spotted spider mite T. urticae were allowed to oviposit for oxe day. The disks were placed on rmioist support tissue in plastic trays covered by lids with aerating holes to prevent condensation. The resulting progeny completed its immature devel- ‘opment on these leaf disks. Four disks (2 cm) of young leaves from each apple tree were settled with one newly emerged female each. These 32 mites per treatment were checked daily for egg production and survival. Eggs were removed daily and leaf disks replaced weekly. In the growth chambers the conditions for plant growth and spider-mite reproduction were 27°C/19°C.(28°C/19°C for the K experiment) with a photoperiod of 16 L:8 Dh and 70% nut. For the determination of the developmental time, approximately 20 ‘gg per treatment of an age difference of not greater than 4 h were monitored, ‘every 4h until emergence to adulthocd. About ten young females of each of ‘these resulting cohorts were killed with CO, and weighed individually on a ‘microbalance (Sartorius, d=0.1 yg). The development studies were con- ducted in growth cabinets at constant temperature of 24°C and a photoper- iod of 16 L:8D. Analyses ‘Chemical analyses were carried out on young eaves, dried for | hat 105°C and subsequently for two days at 60°C. They were analyzed for N, P, K, phe- role compounds, and the four main sugars sorbitol, sucrose, glucose and fructose (summarized in the Result 2s carbohydrates). Nitrogen was deter- mined by a volumetric N autoanalyzer (Heracus), P photometrically (Hita- chi) as phospho-molybdate, and K in an atomic absorption spectrometer (Perkin-Elmer). The sugars were andlyzed enzymatically with UV test kits (Bochringer Mannheim) and subsequent photometry (Kontron), whereas the phenolic compounds were measured as tannin equivalents with the Folin- Ciocalteus reagent (Allen et al., 1974) For the data analysis, only females sufering a natural death were consid cred. From the data on development and reproduction, ifetables were con- structed assuming a general immature mortality of 20% (Herbert, 1981) re- eardless of possible N effects (Suski and Badowska, 1975). For the N ‘experiments, a leaf-N-dependentsex-ratio between 0.73 and 0.77 (Wermelin- get and Delucehi, 1990) was used. For the other experiments, this ratio was sssumed to average 0.75 (Carey and Bradley, 1982). The ife-tables were ana- lyzed according to Southwood (1978), calculating net reproductive rate (Ro) and intrinsic rate of natural inerease (7). This was calculated iteratively by 8 computer program from the equation Berm Lame 1 ‘where fis probability of survival to age class x, and mis age-specific fertility From a the mean generation time (G=lnRo/ra) and doubling time (Dt =In2/rq) of the populations were obtained. The Jackknife technique was used to estimate the variance of these parameters. This computer-intensive ‘method, based on the recombination othe original data is described in more detail by Efron (1982). Examples ofthis technique are provided by Sokal and Rohlf (1981), Meyer et al. (1986) and Krebs (1989), The computer pro- gram calculates the life-table parameters firs from the original sample. From this data set with n observations, each observation if let out in turn and theENVIRONMENT FTES ON TEVRANYCHUS HOST-PANTRUTREEON us parameters are calculated on the remaining n—1 observations. From the life- table parameters thus obtained, the pseudovalues g, for recombination ican becomputed as .=nPp—(n-1)P, Where Pris the parameter (Ro, fm, @ oF DT) of the total sample, and P, the parameters of recombination 1 omitting cbservation i. From the total of pscudovalues for each parameter, the meaa and its standard error can be cal~ culated inthe usual way. ‘Where appropriate, Duncan's multiple-range test was used to separate the RESULTS In all experiments, the concentrations ef the elements N, P and K in the leaves varied according to the nutrient solutions supplied to the plants. The variation of a specific element in the nutrient solution affected the leaf con- centrations, not only ofthis nutrient, bt aartly ofthe other macroelements as well Therefore, nutritional effects onthe spider mites cannot be attributed exclusively toa specific element. All data are summarized in Table 1, while the experiments are discussed individually below. Nitrogen ‘The marked decrease in leaf N concentration was correlated (P-<0.01) with ‘clear eduction of carbohydrates and an equally significant augmentation of the phenolic compounds. Only the controland the two deficiency treatments differed significantly, while within the deficiency treatments and between the contzol and excess treatment (5 N) only minor differences were found. The concentration ofthe single sugars ranged from S to 8% for sorbitol and 1.5 10 44.59% for glucose in all experiments (for details see Wermelinger, 1985). Po- tassium tended to be elevated at N excess supply. The immature develop- ‘mental time and the preoviposition periodof the spider mites were prolonged. in the N-deficiency treatments. Developmental rate, ie. the portion of de- velopment completed each day, was positively correlated with leaf N (Fi. 1}. At the lowest N level, total fecundity beoke down to one-tenth that ofthe contro, the rate of egg production being linearly correlated to leaf N in a range between 1.5% and 3.5% N (Fig. 2). Fecundity was positively correlated ‘with total carbohydrates and negatively with phenolic compounds (cf. Wer- melinger et al., 1985). The temporal pattem of oviposition is depicted in Fig. 3A, While the conteol and the S-N treatment peaked on day 4, the deficiency ‘treatments remained at a more constant, lower level. The drop on day 5 was due to the replacement of the leaf disks tke day before, and the new accom ‘modation and webbing of the spider mites. According to fecundity, female266 A WERMELNOERETAL (Chemin comporiion of spp te aes (CH, sarboysates, PC, phenolic compounds) and ‘sling deveomenta ime, presvpoiion peo, cunt and female weit ofeach wi Pana Koo) Treatment Leaf composition Mie estoy paramees NP Ket re Donel. Proovip candy Longevity Wig (3) (Oh) GDC days) Ga (een) (Ga) Oa) Hivos Sse ioe was ime sa 9 ne ue t Soe ose 9 tos ge oe meses 02 1g Or Re Be 108 mr oe Br Bete 108 YS G9 118 nr has ye sae 1s Le ee AS oo Le Ue Be 6 f Mae aoe ie toe a 9 Mere ae 02 Sem oas La Ie 06 Se ie we Be BP Bos Sah oo Lae Ine noe ane 1 Snae 9 e SN" sar ie 268 oe ans woe tt ise 1 Se Leh 19m toe ioe toe tai Ase Leo Lee [9 toa tor Lo tar 13 Aso LSet Lene Ine 109 Lo me 8 in bse wt expres, fern ar den ign erence (P<001) 5 + Fit. |-Corelaton between developmental reo) o 7 urticaeandN content ofappe eaves weight decreased with increasing N deficiency. With excess N supply, devel~ ‘opmental time was further reduced, but fecundity stayed atthe level of the = N standard. Longevity was unaffectedENVIRONMENT EFFECTS ON TETRANYCHLS HOST LAT UTRIION a6 Fit 2, Covsation between oviposition rte (om) of 7 iar and cantent of pple leaves 91, 3teconidenos intra) Phosphorus ‘While P concentration inthe plant corresponded to the treatments, N con- tent showed an irregular pattern and peaked at mild-deficiency 0.2 P. Phe- nolic compounds and, in contrast to the N experiment, carbohydrates in- creased under P deficiency stress, whereas K did not vary significantly. The ‘evelopment and preaviposition period of the two-spotted spider mite were Slightly prolonged with P deficiency, but fecundity peaked at moderate P shortage where the N concentration was also highest. Correlations of fecundi- ty were found with carbohydrates and phenolic compounds as well as with P content. The pattern of egg production was similar inthe three higher P treat- ments (Fig, 3B), but at 0.04 P, oviposition stayed at a low level. Longevity tended to be incteased by deficiency, and female weight was negatively af fected. Considering all parameters among ie four treatments, strong P defi- ciency differed most clearly from the others, Potassium ‘This experiment showed the least variations between treatments in terms of both plant compounds and mite parameters. At 0.2 K, tendencies to low ‘carbohydrates and phenolic compounds were observed. Corresponding to the faisly homogeneous results of plant contents, spider-mite parameters did not differ strongly either. Only egg produetion and female weight were signifi- ‘cantly higher in the moderate K-deficiency treatment, both being positively correlated with the N content ofthe leaves. The oviposition pattern (Fig. 3C) illustrates the similar reaction to the treatments268 WERNER ETAL Ageldays) Fig. 5. Ovipostion dynamics of. urtoae a fe diferent lve fstogen (A), phosphorus () and potasum (C) spp the hot plats. Lifetable analysis ‘The analysis ofthe lfe-tables ofthe above life-history data on T- urticae is presented in Table 2. Increasing N deficiency was clearly related to a decline ‘of the intrinsic rate of natural increase (rm,). Under unlimiting conditions in ‘terms of space and food availability, a oopulation would grow four times faster in treatment 5 N than in treatment (,04 N, due to faster development and higher egg production. The net reproductive rate (Rp), ie. the number of ‘multiplications in one generation, even differed by a factor of more than I.Life tbe analy for Toon artisan appear at iret eves oN Pad Ks Tresment ra (@ny-9) ny Geen) Baas) : oat soon 453 a5 149" am 006 1 fue tows er 14) ise ae has 2 tise tome tam Ser aah ay nas 0s Ode tno ke ats ae woe 23 Phospons : sou) 41g $38 3s" 20s 2 a8 1 bom 4s 22) 13 202 2 aoe ry Hoo a) ie 202 2@ aon 0s foo ie as ie 203 4 Sam. 3 ous 390 25 40329" 4006 : sho joy $35 302 ae tos 2 Ebi ae hh 3 ae ina 008 shi dee 42 sos jos S01 ‘Baan Ge eb, douing ine means and 9% coin is ‘in tise tat expert, deen tes indicat sigan aiferece (P05). Generation time (G) and doubling time ofthe population were significantly prolonged with N deficiency. The mite pepulation in the 0.04-N treatment needed four times longer to double than that in the S-N treatment ‘The P experiments revealed a more heterogencous pattern. With strong P deficiency, rand Ry were markedly lower, and generatioa time and doubling time were longer than in the other treatments. Moderate deficiency caused, on the one hand, the highest reproductive rate Ro (multiplications per gen- eration), and on the other hand generatios time (strongly influenced by de- velopmental time) was longer than the contol, thus yielding a slightly lower ‘than the two higher P treatments. Only the 0,04-P treatment differed in ry fom the other treatments. Different K supplies resulted in lower rin the excess and strong-deficiency treatments, and in a longer generation time at 0.04 K, However, G and Dr differed only slightly in all treatments, and Ro tended to be elevated with a moderate shortage of K. ‘The best index of a population's performance, i. rq was most strongly related to leaf Nand carbohydrates in the N and P experiments. The most distinct correlation was found between rand the content of phenolic com- pounds inthe leaves with all the treatments pooled (Fig. 4). Above a conoen- {tation of 10% of these secondary metabolites, spider-mite population growthm0 A wERMELNGERETA, Po0%) Fig 4. Relationship tween nase ate of sul inereae (7) of 7; wfc and conten of penile compounds (Fe) inappleeaves (= ~0075-+ODREx G86") ela, 72031). ‘was severely depressed. A tentative, biologically reasonable regression line ‘was calculated, DISCUSSION ‘The response of mite life-history parameters to the nutritional stress cor responds to the pattern generally repocted and documented in the literature reviewed. The life-table analysis, a summarizing assessment of population performance, showed that ry values decreased concomitantly with nutrient supply in the N and P experiments but notin the K treatments, The relatively novel Jackknife technique used for the statistical treatment ofthe life-tables revealed, as far as the N treatments are concerned, significant differences of all calculated life-table parameters between the standard and the two defi ciency treatments. The negative effect of N deficiency on the intrinsic popu lation growth rate (rq) was the result of retarded development (preimaginal ‘development and preoviposition period), diminished egg production and a slightly male-biased sex-atio at lower N supply. Nitrogen was clearly the most important nutrient affecting plant cor;position and lation growth, Net reproductive rate (R) declined deficiency treatment, and the time required for a population to double was {our times longer. In most treatments, development and oviposition of the {wo-spotted spider mite were positively related not only to leaf N but also to the carbohydrate level (primarily sorbitol) and negatively to the phenolic content ‘Toour knowledge, there are only two studies to compare our life-table anal- ysis to, and both investigate 7, urticaeen bean plants, Watson (1964), having two treatments of each of N, P and K, found much smaller differences in Py values between standard and deficiency treatments for both N and P than inENYIRONMED EFFECTS ON TETRENYCUS HOBTFLANTRUTREION an ‘our case, Potassium deficiency did not induce an altered findings. Watson's values fr the standard treatments as determined on beans ‘were consistently lower than ours on apple. The second study was conducted bby Suski and Badowska (1975) with three fertilization levels of each N, P, K. macronutrient. The N experiments showed only a modest increase ofr. with a higher N supply, In contrast to our resuls, excess P supply caused the high- est inerease rates, whereas K shoved the last effects. In both studies, no plant analyses were carried out ‘Some secondary plant metabolite such as phenolic compounds are known to adversely affect pathogens and insect/mite populations (Ishaays, 1986; ale, 1988). In the present study, total phenolics were strongly negatively related to spider-mite development, oviposition (cf Larson and Berry, 1984) and population development, as indicated by rq. On the other hand, negative correlations were found between phenols and nitrogen ot sugar content. ‘Therefore itis not possible from this study to pinpoint a singe factor as the ‘true cause controlling the development of spider-mite populations. In fact, the balance between the adverse effects of phenolic compounds and the ben- ficial effects of amino acids and carbohydrates is suggested as the determin- ing factor in the performance of a population. This may be illustrated by the experiment: leaves with sight P deficiency have a somewhat lower concen- tration of phenolic compounds and more N and carbohydrates than in the standard treatment. This coincides with the fecundity peak. With severe de- ficiency, where fecundity suddenly drops, phenolic compounds are markedly higher and N is lower, thus overriding the favorable effect of still elevated carbohydrates. Similar balances were found between temperature and food ‘quality under water stress (Wermelinger etal, 1990b). ‘A noteworthy aspect of all our experiments was the increase of variation with stress (ef. Table 2). This phenomenon is best evidenced by the parame- ters affected by development (rm, G, DT) in the N experiment, where stress ‘was apparently most severe. Although the mean values of eg. fy inthe N {deficiency treatments were considerably smaller than with sufficient N sup- ply, their variation was much higher. Under unfavorable conditions, the = hetic heterogeneity of population, in tems of individual fitness of the fe- males, seems to become more evident than in an optimal environment. Professor A.P. Gutierrez, U.C. Berkeley, kindly provided an algorithm on ‘which our computer program for life-table analysis was based. Thanks are due to Mrs. M, Graf for reviewing the manuscript and to an unknown reviewer for his helpful suggestions.REFERENCES Abou-Awad, BA, 1981, of trogen and manganese eration on he tetany mite Teranehus arabicus Ata, and sung det on ld of eoton pans Bul Soe. Ento- sno Egypt, 93: 95-98 Alles, SE, Grimshaw, HLM. Parkinson, JA. and Quarmby, C. (Editors) 1974 Chemical ‘Atalysisof Ecological Materials. Blackwell, for Archer, Byoum ED. Je and Onken, AB, 1988. Abundance of Banks grass mies (Aca: “Tevanjchidae) on corn abd sorghum feilizd with ifeent rates of nope and phospho us J. Eeon Entomol 1300-303, Cate, and Brae, 1 W., 1982 Developmental rates, vt sebeuls, se ais, and ite ‘ables for Teramys unica, T. wkestan and T. paca (Acasa: Tetranyenie) On fatto, Acaolgia, 23: 333-345. Dale, D. 1988 Pan’-mediate fects of sol mineral stresses on insects. In EA. Heivichs (Eto), Plant Sree nsct Interactions, ley de Sons NewYork, p. 35-110. tron, B, 1982 The Joka the Bootstrap ard ater sampling plans Soca fr Indus 1nd Applied Mathematics, Philadephia, PA. 92 pp. Epstein, (Editor), 1972 Mineral Norn of Plants: Peiacpes and Pespecives Wiley, ‘New York. ‘riasche, Wolfgang H., Res. and hide S, 1980. Untersuchungen zu den Urachen ‘ortesbedinger Befalsuterschiede von Apesamen mit Olgomchus ulmi Koth. Arch. Phytopathol. Pnanzenschuta 16: 198-198 ‘ber, 198 Biloy ie ables, nd inate apa for increase of he wospatted si ‘er te, eranyehusuricae(AcaronTetaayctae). Ca. Entoml, 113-371-378 ‘Mhaaya, I, 1986. Nuional and alllochemi nsec-lant interactions Fang to digestion sh food inte some exmples In} Miler pT A. Mile (Ears), Iaet Pan le teractions Spange, New York, pp 191-225 Jackson, PR and Hustr, PE. 1983, Ee of nitrogen frie evel on development and ‘sopultons af he psc ef scorch mite (Acar Tetanjhidse). J. Boon Estomel, 6 Bias Jest L,Sok, ZW. and ado. Cb. 1979. Food quality infisences on pier ‘ate population. In: 0. Rodeigucr (Editar), Recent Advances in Acarlogy, Vol L ACE demi Press, New York, pp. 189196, ‘Kondo, A and Takai, A 1982, Eft of bo quality om se ati in the twosptted spider ie, Teno ileae Koch (Acris: Taranyhida)-Jpa-1 Appl Entomol 20a, 26 200-202, ‘Krebs, C1, 1989, Ecologia! Methodology Harper & Row, New York, 654 pp. Lane. and Schuster MP, 1981, Condeasd tannins of oton eaves Phytochemistry, 2 "5-427. ‘Lason, KC and Bers, RE, 1984 Influence of peppermint phonois and monoterpenes o8 Swope spider mite (Acar Tetanjeice)- Environ. Entomol, 13:282-285, Matson. W4. Je, 1980. Herbivoryinrelation plat nitopen content Anau. Rev Ecol Sst, 11 19-16 Melrs, WK aod Props, SE, 1983 Metso ftize level fetity balance, and soil mois ‘ure onthe interaction of to-poted spider mites (Acaring:Tetraychie) with radish plant: Environ Entomol, 1212391246 Meyer 1S, lngesol, CG, MeDonuld, LL and Boyer, Ms, 1986 Estimating uncertainty in ‘dopalaton promt ates Jakke vs, Bootstrap echnigues, Ecology 6: 1136-1166 ering, TM, Archer, TL, Johnson, 1 W. and lips 1.M, 1982 Evauaion of several pain ‘orahum characteris for resstance io the Banks asst. J. Econ Entomal. 75:257- 26.VIRONMENT EFFECTS ON TETRANCHUS: OST PLANTRUTREION a Pertng, TM, Archer, TL, Krieg, D1. and Johnson, JW, 1983. Relationships between the Tank rst ie (Acaridrmes Tetranyhidee ad piysologcl changes of mating rain sorghum. Environ Extomel, 12: 1034-1098. Rajaram, 1A and Lang Hos, 1975, Eel ofborn auton on intensity of ed spider ie tack ool palm sevigs Exp. Agric, 1159-63, Sharma, Band Pande, V.D, 1986 study of elaionshi between the population of Tera Shu necledoicus Ande (Acarna Tetanychae) and external characeristsof cua Iitleserang their NPK contents 1. Ad Zool, 2-45, Sola, RR and Rol, 98 Biometry. WH. Fran, San Franco, 859 pp, Southood, TRE (Editor), 1078 Ecological Meihcs, With Particular Reference othe Study ‘of nsec Populations. Chapman & Hal, New Yok, 328 pp Sere, M, Ree Sand Coldenberg 1975. sf of chemical deficiencies in tren Toluions a the reproduction of Acer sheldon (Acsrins, Erophidse). Bul. Entomol es 65: 433-42. Sushi, ZW. and Badowsa, T1975. fet of hos pant mutton on the population of teow ‘sped spider mite, Tevanjouswnicae Koch (Acina,Tetanychida) Exo PO, 23: 185- 0 Tull, U, 1971. Fee and bound amin acids ofthe host plant species nd varius eter ‘tetmens affecting te fecundity ofthe twospoted spider mite Tevanjehus wae Koch (Aesriny, Terayehide), Ane. Entomol Fenn 7: 155-168, ‘on der Geet, LPS, 1985. Studieson artical is fo spider ites: W. Helle and MW, Sabels (Editors, Spider Mites, Their Biology, Natural Enemies and Conta, Vo. A. Ese- ‘er, Arsterdam, pp 385-390 Van de Vie, Mand Dever, P1979. Nitopen fertization of fait ree and its consequences “or the development of Pononshus wn populations and the grow of rat es. 1.6 Rodrigues (Eaton, Recent AdvanesinAcarley Vol Academie Pres New York, pp. ra ‘Vande Vii, M, MeMurry, JA. and Haller, CE, 1972, Bslgyoftetranychi mites and hee natral nee avon I Biology, coi, and pest tats, nd hot pant lations oftetranyehis Higa, 41: 343-42, ‘Watson, TF, 1964, Inuence of bos plat condition on population increas of Teranychus clas (Linnaeus) (Acafon: Tevanschidat) Higardis 38:273-322. \Wermlings, 1985. Enflos ds Errungaratandes der Wispanz auf den Bef durch “ie Gemeine Spinamile (Teranotusurtcae Koch). PuD. thesis, ETH Zieh/CH, 99 p. ‘Wermeinger,B, 1989: Repiration ofthe wo spot spder mite as etd by laf nirogn “APA Eaton 18: 208-212 \Wermelinger Band Dees V, 1990, The effet fsx ation the mutipiation of he wo- ‘oted spider mite a affeced by ef mrogen. Ep. Appl AcaOL9: 11-18 \Wermclinge, Or, and Deluch,V 1985 Eis of hot plan irogenFeriization ‘onthe biology of the mo-spoted spider mite, Teranjhus wie Eaton Exp. Appl, 3: mh, Wermelinge,B, Boumgsrar, J, Zahner, Ph and Deluchi, 19903, Environmental actors Mfecting the ie blesof Tevanychas nce Koc (Acar), 1. Temperature. Mit, Sche Enlomo Ger 63 55-62, Wermelinge, B. Schnider, F, Oe, 1. and Baungirnr, 2, 1990, Environment factors fing the fe tables of Tenschus rae Koen (Acarina). IL Host plant male tes. ‘Mit See. Entomol, Ge. 3347-359. Wisoa, LT. Sila, 1M, Hofmann, MLP, Faery, DL. and Ruiz, My 1988 Le ni ‘wogen and position in relation to population parameters of Pace spe ie, Teranchas paces (Acar Tetanyhidae) on grapes Enon Entomol, 17 964-968,‘Wressc, D.L snd Young, SS. 1978 fect of density and host aly on rt of develop "nent survivorship, and sex rae the caine spider mite Environ. Entomol, 7 499- ‘Yenineh, 1S. Guserer AP. and Herren, HR, 1999, Dynamics of Mononycholu fanaoa (Acar Terayehida) in Affe experimental evidence of temperature and os plane ‘ees on popalation growth rates Environ Eom, 18: 633-640.